Research coverage

What the research covers

A published research snapshot. Sources, deployment stages and review dates appear on each case. Editorial research is broad discovery, not exhaustive worldwide coverage or continuous monitoring.

1002 researched workflows · 1059 catalog records

Latest research round: 2026-11-08. Each case retains its own review date.

Regions searched: Asia Pacific, Singapore, Australia, New Zealand, Japan, China, Hong Kong, India, South Korea, Indonesia, Malaysia, Thailand, Philippines, Vietnam, North America, Europe, Global, Middle East, Latin America, Africa, United Kingdom, France, Germany, Spain, Netherlands, Belgium, Nordics, Switzerland, Central and Eastern Europe, Italy, Portugal, Greece, Asia-Pacific, Middle East and North Africa, Asia Pacific (China, Hong Kong, Singapore, Japan, South Korea, India, Australia), Latin America (Brazil), Africa (South Africa), Latin America (Brazil/Mexico), Asia Pacific (India, Hong Kong, Singapore, Japan, Australia), Middle East (Saudi Arabia, UAE), United States.

Coverage gaps and access limits

This is a broad public-evidence map, not exhaustive worldwide enumeration. Undisclosed internal deployments, unsearched subsidiaries and every small institution are outside the observable boundary; search absence is not evidence of non-use.
North America: community and regional banks beyond the named cohorts remain unevenly covered. Credit-union breadth is concentrated in Zest AI and Eltropy public case libraries, which reflects publishing availability rather than market share.
Europe: smaller German savings/cooperative banks, Baltic/Balkan domestic institutions, Ireland and smaller Iberian banks remain undercovered. French and other regional-language searches were not systematic; several issuer or report retrievals failed.
Asia Pacific: mainland China, New Zealand and several Korean and Southeast Asian institutions remain undercovered. Systematic native-language searches were not completed for every institution; ICBC and several Korean primary documents remain inaccessible leads.
Latin America: Brazil supplies most qualified banking depth. Non-Brazil banking, Central America, the Caribbean, Andean institutions, local insurance and private funds need additional targeted coverage.
Africa and Middle East: evidence is concentrated in South Africa and selected named operators. North Africa, Francophone Africa, West Africa beyond Nigeria and regional smaller institutions remain thin; no systematic Arabic, Hebrew or French-language cohort pass was completed.
Private capital was searched in the final round, but coverage is concentrated in North American/European managers. Small private-credit firms, specialist fund administration and standalone Africa/Latin America/Middle East manager cohorts remain undercovered. KKR filing access and Brookfield manager-operated workflow evidence remain unresolved.
Digital assets: Coinbase and selected analytics operators supply concrete workflows; custody-specific controls remain unconfirmed. BitGo filing access, Kraken assistant evidence and named-client attribution for several analytics products remain unresolved.
Internal functions: ledger-close automation, continuous financial-control testing, access-provisioning governance and staffing optimization remain thin. Cross-application reconciliation must not be relabeled general-ledger close; asset-manager-specific HR and cyber evidence is still limited.
Insurance: life/health actuarial modeling, reserving, optimization and internal cyber evidence remain thinner than service/document workflows. Regional brokers, pensions outside the main European/South African examples and Japanese-language insurer coverage need further depth.
Markets and liquidity: collateral and settlement-risk examples are documented, but evidence does not establish a broad population of autonomous settlement engines, liquidity optimizers or independently verified AI investment alpha. Several institution/initiative identities remain lead-level until resolved.
Financial infrastructure: named-client AI use is thinner outside financial crime. Generic AI-enabled platforms, modernized cores, rule-only workflows and anonymous customer cases do not establish a named operator's specific AI deployment.
Public financial authorities: EBA, OSFI, BOJ/JFSA, RBNZ and several African authorities mostly yielded governance or broad-program material; some RBIH, BCB, CBUAE and SEBI workflow evidence remains indexed or incomplete. BIS experiments and future plans are not production deployments.
Non-FS transfer coverage is concentrated in North American/European organizations, with selected Asian and global-public examples. Industrial, health, legal and service-operations lanes did not complete exhaustive Africa, Latin America or Middle East cohorts; worldwide company footprints do not establish regional research saturation.
Some concrete source leads remain blocked, indexed-only, oversized or video-only. They stay visible with access reasons and require targeted retrieval; neither credible headlines nor elapsed launch targets establish a completed deployment.
Historical and undated operating evidence is common. Review date is not launch date, and a production claim at its source date does not prove continuous operation or unchanged scope through October 2026.
Independent outcomes, full implementation/review costs, material-error rates and comparable quality denominators are usually undisclosed. Workflows without ROI metrics remain eligible; issuer/vendor value claims, forecasts and model benchmarks are not independently verified savings.
Practitioner evidence is a purposive sample of public Reddit discussions and X conversations. Identities, employment and underlying events are mostly unverified, ranked replies are nonrepresentative, and anecdotes supply evaluation hypotheses rather than prevalence or causal estimates.
Bounded English, French and Arabic discovery, not a census. Thirty-two actual search strings are logged. Extra retrieval calls beyond the researcher's approximate 15-call guidance resolved image-only PDFs and checked existing leads; no saturation claim.
Qualified African additions are concentrated in South Africa. Francophone Africa, West/East Africa, North Africa beyond the Morocco lead and discrete Islamic-finance workflows remain undercovered; no inference of non-adoption.
Arabic searches were executed, but no new Arabic-language full-text primary deployment passed inclusion. Bank ABC's /ar/ URL hosts an English report.
Ajman Bank/SAS partnership, training, hiring, broad strategy and unnamed Moroccan-bank vendor cases were not counted as deployments. No new standalone Islamic-finance use case was verified beyond baseline institutions.
Sanlam's issuer homepage timed out. Its actual 96-page issuer-authored report was retrieved in memory from the FinancialFilings PDF mirror; complete relevant pages were directly extracted. Correct location: printed p.74 / PDF p.75. The web HTML transcription mislabeled the passage p.75 and contained OCR artifacts. AI performance claims are not independently audited by virtue of appearing in this report.
FAB case PDFs expose only titles to text extraction. Complete cyber, trade and compliance pages were rendered in memory and visually read. Web screenshots produced no inspectable images; oversized batched image output was truncated, then individual renders succeeded. No source binaries or render files were saved.
ADCB annual report exceeded web size limits and direct retrieval returned HTTP 403; no bypass attempted. Included ADCB claims rely on its accessible 2026-01-29 earnings release, p.5. Annual-report-only architecture and AML assertions were not promoted.
Bank ABC direct retrieval returned HTTP 403; the ordinary web tool retrieved the public full PDF successfully. Relevant complete page text was inspected. Publication date remains unknown; chairman statement and financial-statement authorization date are 2026-02-08.
All stages are source-supported historical states, not independent observation of current operation on the review date. Most sources lack model evaluations, failure rates and outcome denominators.
Simple AI is a secondary-only French-language pilot lead. Official Simple homepage was fetched but contained neither AI nor intelligence in extracted text; this is lack of corroboration, not counterproof.
Compared all 21 existing *.candidates.json files (950 candidates) by organization aliases and workflow, and searched baseline-catalog.json. This is bounded duplicate screening, not a fresh audit of all application records.
Only one qualifying Oct 27-29 issuer deployment release was established: DOCOMO SMTB Net Bank, dated Oct 28. Older missed deployments remain eligible. China Oct 29 Nanjing Bank result had generic AI-scale-up language, no sufficiently specific named workflow, and its full page was inaccessible.
Korean issuer PDFs were blocked by size/timeout: Samsung Fire 2024 integrated report exceeded the web fetch limit; Hanwha Life 2025 sustainability report timed out. Samsung underwriting and Hanwha Financial Data Agent remain leads, not production inclusions.
BOCHK annual report exceeded the fetch-size limit; its Frontline GPT workflow was instead verified in the bank-authored case entry in the official 2025 Hong Kong ICT Awards booklet, with issuer annual-report search text as corroboration.
No claim of complete China/Japan/Korea/Hong Kong coverage. Chinese rural/regional banks and Korean brokers remain undercovered. No ROI, error-rate improvement, or current 2026 live status inferred from older releases.
No direct adverse evidence was found for the eight included workflows in this bounded search. Absence of counterevidence does not establish effectiveness, safety or continuing operation.
No qualifying new deployment published October 27–29 was verified. This is predominantly an older-missed-case update, not proof of no activity.
Regional coverage: CEE is concentrated in Polish insurer PZU; Nordic evidence is concentrated in DNB. Domestic Baltic, Balkan, Finnish and Danish institutions remain thin.
Treasury searches returned Nordea/SEB future-oriented discussion, not a named live treasury AI workflow; no production treasury case included.
KBC 2025 sustainability PDFs on both kbc.com and wcmassets.kbc.be exceeded the browser tool's 10 MiB fetch limit. Indexed passages are supporting leads only; KBC inclusion rests on fully fetched 2023 annual report.
Santander Consumer Norway's AI credit/pricing statement was indexed but full page yielded only a shell; not included.
AI-specific job advertisements and conference agendas were not accepted as deployment evidence. PZU's keyword-based subrogation robot was not counted as AI merely because it appeared near AI claims.
BCF evidence is named-client vendor reporting, and DNB cyber evidence is a direct practitioner interview in trade press, not an issuer filing.
No independent model-quality assessment or causal ROI evidence collected. Historical stage statements are not silently advanced to 2026.
Baseline comparison parsed all *.candidates.json and checked institutional/workflow names; targeted rg found no Kamuno, Samoobsługa, StiPPLE, Agro Lab, bodily-injury, NN Group or Freiburger matches.
One bounded verification fetch beyond the research guide's typical 15-call target was used to open UKB's newly found annual report and resolve Kamuno beta versus full operations.
13 web tool calls completed, including four batched search calls with 15 unique query strings; remaining calls fetched or inspected primary texts. No HTTP 429 or authentication/access bypass occurred. No connector fallback was required.
Baseline compared by parsing all *.candidates.json institution/title fields, then reviewing relevant software-enterprise records and searching exact workflow names across all baseline candidates.
New Oct 27-29 deployed evidence was strongest for Meta PrivacyLib (Oct 27) and OpenAI monitoring (canonical header Oct 29; initial search index Oct 28). No qualifying equally recent Anthropic, DeepMind, Microsoft or AWS-internal deployment was established; older missed disclosures are explicitly dated.
Source reporting is generally first-party. No independently audited FS ROI, production bank accuracy or loss-reduction result should be inferred from these transferable cases.
METR October 22 red-team report was located and opened, but the tool did not expose usable report body in returned output. It is not used as independent substantiation; OpenAI's own incident report supplies the documented counterevidence.
Amazon RL gyms are historical January research operations, not asserted active October programs. AWS WGU is a named customer deployment, not an AWS-internal deployment.
DeepMind scam model rollout and Anthropic purchase-review implementation remain leads; missing operating detail is not filled with assumption.
Protocol read first from the assigned researcher.md. Read CLAUDE.md and opening project progress/lesson context; no application, database, runtime, or project-document changes made. Only this lane's two research outputs are written.
Baseline comparison: parsed health_life_sciences.candidates.json and legal_professional_services.candidates.json, then rg searched all *.candidates.json in Global financial services AI use cases for Dinsmore, Garfield, Ashurst, UCLA, Wisconsin, PSI CRO, Evidently, Legatics, Stewart Law, Cofactor and Memorial Health. No matches. Novelty means absent from these candidate files, not globally novel technology.
Freshness is source-specific: three included records are October 28, 2026 reports (PSI, Stewart, Everlaw early access). Older missed cases retain their original as-of dates. No independently verified October 29 institutional deployment found in this bounded pass.
Access log: PMC12768499 and PMC12858090 returned browser/reCAPTCHA challenges. NEJM AI DOI 10.1056/AIoa2500945 and the medRxiv playbook PDF returned 403. No challenge bypass attempted. UCLA and Wisconsin institutional full-text releases support the included operational cases; exact Wisconsin trial statistics beyond that release remain a verification lead.
Access log: National Law Review Stewart release returned 403; accessible EIN Presswire copy of the firm's release supplied full text. Its framework link resolves to a lawyer bio, which also returned 403; this is not an independently obtained written framework or podcast transcript.
No HTTP 429 was encountered in this lane. The web search tool remained functional, so no connector fallback was necessary. Searches and dependent full-text opens were batched; blocked sources were not repeatedly hammered.
Memorial case study is undated and linked to mhsystem.org. A secondary search result described Florida; that location was not adopted. Exact deployment date, denominator, and causal outcome evaluation remain absent.
University of Iowa training page is undated. Its vendor's July 18, 2025 release supplies the September 2024 rollout chronology. No claim that Iowa is one of Evidently Anywhere's unnamed October 2026 deployers.
Ashurst is retained as a counterevidence/evaluation lead, not a production client deployment: the 2023-24 trials expressly used public rather than client data. Exact report publication day not established.
Third-party and anonymous appeal case studies were not promoted to named institutional deployments. No healthcare ROI, clinical efficacy, reimbursement gains, or vendor market-wide performance was imported into FS projections.
General limitations: issuer/vendor testimony can substantiate reported workflow use but rarely supplies independent error rates, denominator, total cost, or controlled causal savings. Test/kill criteria below are proposed FS experiments, not observed institutional acceptance thresholds.
Research used the evidence distinction in aluda-expert-review/SKILL.md; no numerical persona scores or real-reviewer claims were made.
No new, sufficiently specific deployed case first announced on 2026-10-27 through 2026-10-29 was verified in this bounded search. C.H. Robinson is an October 26 missed gap; the other admitted cases are older or undated evidence.
Recent syndication is not event timing: October 29 CEVA/Sereact coverage traces to the October 21 primary release at https://www.globenewswire.com/news-release/2026/10/21/3385409/0/en/sereact-nvidia-nebius-ceva-logistics-and-zalando-close-the-training-loop-for-robotic-returns-handling.html?trk=public_post_comment-text. That release describes a robot-policy simulation/training/evaluation loop and an initial Zalando returns deployment. Not included as an FS operating case: transfer from physical robot policy learning to regulated insurance assessment would need more direct evidence.
Batteries Plus discovery coverage appeared dated October 26, but its retrieved original product release is August 31. Expected end-September completion remains unverified.
Exact publication/status dates are absent on the Exelon, OluKai, SiriusXM and Kraken customer pages. Historical dated corroboration is used where available; no present-day enterprise-wide operation inferred from crawl dates.
Duke Energy searches returned inspection programs and older planned AI work, not adequate full-text proof of the specific AI deployment sought; Exelon closed the named utility-inspection gap instead.
No independent audits, bank pilots, causal ROI studies or regulator endorsements of these transfer proposals were found. All FS application and test gates are analyst inferences; source-sector ROI is not imported.
Geography remains US/Europe-heavy; C.H. Robinson includes China ocean bookings and other-market quoting, but this is not a comprehensive global industrial or retail sweep.
Public web search/open was available throughout. No 429, paywall bypass, authenticated browser use or connector fallback was needed. Baseline comparison covered the two assigned lane lists in full at title/stage level and name/initiative searches across all 21 candidate JSON files.
Direct full-text verification failed for the Aon/SAGE October 29 release and Swiss Re pages; secondary search-index detail retained as leads only. Connector fallback produced the same inaccessible/empty results. Ordinary HTTP retrieval failed DNS; no bypass attempted.
No newly dated October 28 named-carrier operational case was validated. October 27 Guidewire is supplier availability/roadmap, not deployment.
No verified insurer-wide AI statutory reserving, IFRS 17 liability valuation or reinsurance capital-model deployment established. Gallagher Bassett is claim-level reserve guidance; WTW is product-only.
Named insurer-level AI failure or loss causally attributable to deployment was not validated. EIOPA supplies operating-burden and maturity counterevidence, not a failure count.
Aria Advisory/Akur8 selection describes deterministic/stochastic reserving; supplier AI branding alone does not establish that the deployed reserving method uses AI. Frank Winston Crum release is selection/future support and was not promoted.
Older releases are retained with source dates; no present-day deployment reconfirmation for 2025 operational accounts.
Baseline comparison covered life_health_insurance and property_insurance candidate titles plus cross-file title/institution matches for target entities/workflows. Baseline source bodies not all exhaustively reread.
No verified October 27–29 deployment change found; findings are older missed cases, not newly launched claims.
Chile issuer PDFs for Banco de Chile 2024 (11.4 MB) and Bci 2024 (19.7 MB) exceeded web fetch limits. No authentication bypass attempted. A later primary Banco de Chile CEO interview supplies FAN-credit evidence.
BAC 2023 PDF (17.7 MB) and Banco Industrial 2023 PDF (11.7 MB) exceeded fetch limits; BAC remains a gap. Banco Industrial was verified using its 2020 official post and readable 2025 issuer-report embed.
Canonical Issuu URL could not be opened. Readable public embed identifies Corporación Bi as publisher. Banco do Nordeste 2025 full issuer report was verified through an Investidor10 mirror, not its original host.
BCP 2025 report was partly retrievable, then returned status errors. Generic organizational AI statements did not establish a sufficiently specific workflow.
RIMAC, Grupo SURA, BancoEstado, Banorte, broader Caribbean and non-Guatemalan Central America remain undercovered; generic partnerships, awareness articles and job advertisements were not treated as deployments.
Nequi technical evidence is a vendor-authored English case; no equivalent full issuer Spanish technical account was established. Exact source publication and deployment-start dates are unknown.
No loss-adjusted credit ROI, audited accuracy or causal financial outcome was independently verified.
No independently qualified new production deployment dated October 27-29 was found in this bounded pass. DMCU was a prior-day selection republished October 27.
Treasury remains weaker than collections and internal assistants: Regions lockbox lacks a disclosed technical AI mechanism; M&T cash-application article was generic product guidance without a named deployed initiative.
Canada regional/credit-union coverage remains thin beyond First West. Its annual report timed out; later workflows rely on a full-text vendor customer story with named executives.
CU Times Kitsap/OnPath/Heartland article was indexed but its fetched body was absent. No paywall/authentication bypass attempted.
PNC October 27 coverage discussed strategy, spending and expected savings; it was not treated as a new deployed workflow.
Valley prospecting/fraud production is attributed to its COO in July 10 Banking Dive reporting, but the technical AI mechanism was not identified; not promoted.
Servus F5 deployment was found, but reviewed customer passages described broader ML/AI exploration rather than explicitly identifying the deployed bot-defense AI mechanism; not promoted.
Customers Bank-Palantir October 15 announcement used future deployment language for deposit onboarding; not merged into existing coding deployment.
Queries ran through web search without a 429 in this lane. Only the assigned two research files were written; app/database/runtime and project documentation were untouched.
Novelty checked across all baseline *.candidates.json by institution and workflow. Existing WSECU underwriting, Huntington fraud lead and Regions recruiting lead remain distinct.
Bounded discovery, not exhaustive coverage. Most evidence is older missed material; no independently verified new Oct 27–29 deployment is claimed.
Recent coverage often re-dates older events: Visa pilot was Oct 20, not Oct 27; Raise appeared in Oct 27 coverage but an older vendor edition is dated Mar 7, 2025.
J.P. Morgan Prysmian page and Treasury Insights page returned HTTP 403. Normal search-connector fallback for Prysmian also returned 403. Search text is retained only as lead; no bypass attempted.
No web 429 was encountered in this lane. Search and opens were batched serially; connector fallback used only for the blocked J.P. Morgan source.
Baseline comparison read candidate institution/title/decisions and searched all *.candidates.json for exact institution/product names; Ant Falcon is an existing provider workflow, so Capital A is adopter-level novelty and may belong as evidence on that record.
Personio, Sodexo, Capital A and Raise are explicitly transferable corporate/merchant cases, not new bank deployments.
No new fully verified merchant dispute-representation workflow was found beyond baseline Stripe/Visa; Raise is chargeback prevention.
Digital assets: Sygnum is a controlled live pilot, not client production; RockWallet lacks explicit AI attribution; Ripple's customer list does not establish GSmart activation.
No vendor-reported outcome is treated as audited ROI or causally attributable to AI alone. Banking regulation and consumer-debt applicability are not inferred from B2B corporate cases.
Konica Minolta/HighRadius, Phaidon/Sidetrade, Domino's/Amtrak/J.P. Morgan surfaced as additional leads but were not promoted without bounded full-text case verification.
Some initial literal site.domain queries were broader text queries rather than site: restrictions; inclusions rely on opened primary pages, not assumed domain filtering.
Bounded discovery, not a global census. Most included government implementation evidence is UK; geography was not padded.
Found two new 27 October sources (Defra launch and Anthropic security research); no qualifying new 28–29 October deployment established. Older missed cases are labelled by their actual publication/update dates.
No additional scientific discovery or procurement decision deployment with sufficiently specific primary evidence was established. Laurium is policy research analytics; PSFA supports risk-assessment preparation, not automated procurement award.
Public ATRS records are operator reports; detailed metrics and independent evaluations are often absent. Publication date is not assumed to be deployment date.
All financial-services applications and test gates are inferences. No non-FS outcome or ROI has been imported as an FS benefit.
Ofsted survey analysis remains a lead because live use is not established; formal verification remains a research lead; the retired Ofsted writer is excluded from active counts.
No authentication bypass, application/runtime/database edits, benchmark execution or outreach performed.
No new operating-AI use case dated October 27–29 was validated in this bounded lane. Results were dominated by guidance, conferences and AI-security policy. This is not evidence that none exist.
Bank-owned AI assisting model validation remains thin. Search surfaced governance commentary and regulatory guidance rather than a named operational validation workflow; none promoted.
Four Bank of Italy entries and two CONSOB entries share official sources. This is workflow coverage, not six independent institutional deployments. OECD Table 2.2 is institution-supplied; PDF and HTML are the same evidence.
Do not equate the 2026 six-tool production total with automatic production status for every table item. CONSOB is explicitly prototype; Banca d’Italia complaint tools have mixed testing wording and were not added.
BaFin annual report returned HTTP 403. No authentication bypass attempted. Vectra customer PDFs failed in both built-in web and connected search-service fallback.
mLeasing surfaced in a primary PDF index (2021 version); full PDF inaccessible at both linked CDN and search-discovered public asset URL. Excluded from findings rather than counting a fourth weak lead.
DNB EC Technical Support Instrument project page (programme year 2021) describes planned MVPs and future deployment; current live workflow status and specific product identity not established.
Metrics deliberately withheld: no independent ROI; Bank of Italy timing graphic is extraction-only simulation; board charts hypothetical; NAB Windows benefits not Security Copilot outcomes; Blackstone directory metric descriptions differ.
Novelty checked by reading baseline regulatory/internal-function candidates and rg searches across all *.candidates.json for ACPR/LUCIA/Veridic/BaFin/ALMA/Banca d’Italia/CONSOB/ownership/board and NAB Security Copilot/Vectra/Hibank/Blackstone cyber identifiers. This is a candidate-file dedupe, not application-database validation.
Search tools did not return HTTP 429 in this lane. Connected fallback was used for failed public PDF retrieval, with no success.
Compared all 21 baseline *.candidates.json files: 950 candidate rows, plus the separate emerging-markets leads field. Workflow, institution, tool-name and source-overlap checks performed. No app/database runtime changes.
One verified publication in October 27–29: MarketAxess CP+ PT, 2026-10-27. Other findings are older missed sources or undated issuer pages; do not present them as October launches.
No strong primary evidence found here for a named CCP using learned models to set live margin, manage defaults, or clear transactions. Standard margin calculators, optimization programs and cloud partnerships do not by themselves establish AI.
Danske differential-ML XVA deployment remains a lead outside findings: primary deployment verification not obtained. Risk.net 2020 account describes testing and future production, not proof of present use.
State Street 2023 annual-report PDF was indexed but fetch exceeded 19 MB tool content limit; not used as full-text support. Accessible issuer article independently supports reconciliation.
Crédit Agricole PDF text was accessible; one page screenshot returned cache-miss. No blocked or index-only source supports an include decision.
OCC 2025 GenAI release verifies developer testing/release-note workflows, not AI clearing decisions; omitted from this lane. Broadridge anonymous bank reconciliation case was not promoted to a named-bank finding.
SG's baseline mentions a broader existing registry AI Trading Analytics item outside this candidate corpus. Trading Assistant is evidence enrichment until deduplication against that item. The PDF cover says February 2025 despite URL folder 2024-09.
No performance or ROI figures imported; experiment results, vendor model comparisons, prediction accuracy, and realized economic gains remain separate.
Compared wealth_asset_management and private_capital_gap candidates plus cross-lane literal searches for named additions. Baseline review date was 2026-10-26; novelty means missing/previously unvalidated evidence, not necessarily recent launch.
No fresh Oct27–29 named private-credit monitoring deployment established. Citco's 26 October pilot is the nearest fresh operational evidence; Capita reporting on 27 October is a strong negative case.
APG 2025 annual report alternate assetmanagement.apg.nl URL returned 404. Existing APG Genie lead remains unpromoted. APG next-best-question already exists in life/health lane.
TPR 29 October guidance retrieved only as index text; both web and connected fallback failed. Citco 29 October middle-office article cache-missed; do not infer AI from title.
Canoe AI and product-detail pages returned 403. Public Why Canoe page retrieved and explicitly describes AI/ML; GCM interview supplies deployment evidence. Gated Lake Street and Prime Buchholz case studies were not bypassed, and not added as strong cases.
Cardano brochure redirected to Mercer homepage. Citrus inclusion rests on full-text December 2024 CFA report, not redirected brochure. Current continuation not checked.
Recent searches were heavily contaminated by investments in AI companies and data-centre debt. These are not evidence of operational AI and were excluded.
73 Strings unnamed top-ten client and logo list are not sufficient to identify additional customers. No private corporate identity was inferred.
Alvarez & Marsal primary 73 Strings partnership PDF was retrieved and describes implemented valuation support, but omitted to avoid excess adjacent professional-service cases; URL: https://www.alvarezandmarsal.com/sites/default/files/2024-11/9383_A&M_446837_%20VAL-US_73%20Strings%20Platform%20Launch%20Social_2PP_version_LC_STG02.pdf
Independent outcome evaluations, error rates, material valuation challenges and savings net of AI/platform costs were not found. Customer/vendor metrics are not ROI.
SS&C release also covers ConnectIQ, Black Diamond and Genesis; no named production adopter established in this pass. Announcement evidence should not be counted as deployed cases.
Public source retrieval only. No app database, runtime, project documentation, authentication or email changes.
47 search queries plus source opens/clicks/finds; bounded discovery is not comprehensive worldwide coverage.
Searches in English, Portuguese, Spanish, French, German, Japanese, Korean and Chinese. Korean/Chinese/India searches did not establish additional qualifying primary deployments in this bounded pass.
African RexAtlas lead remains held; no verified new African banking deployment accepted.
MENA Mashreq/Clari5 fraud-platform discovery sent to parent for fraud-lane adjudication, not duplicated here; Emirates NBD event participation alone is not deployment evidence.
HSBCnio November secondary coverage recycles the issuer's 2026-09-29 launch and is not a new refresh-window case. Deutsche Bank AI Academy is training infrastructure, not itself an AI workflow deployment.
US major-bank searches yielded no qualifying new primary workflow in the bounded pass beyond the Foundation Mortgage launch.
Santander/Noxus and ING mortgage findings remain leads pending primary corroboration. No provider model calls or external communications used.
NAB PDF screenshot failed with cache miss; textual primary extraction clearly supplies memorandum-preparation measure and date. Westpac filing accessed as an ASX issuer document via YourIR mirror.
No newly found primary material updates to an existing workflow other than Nubank were established; independent adjudication should still inspect institution and initiative overlap.
45 search queries plus direct page opens and within-page searches; bounded discovery, not exhaustive global coverage.
No fresh scientific deployment accepted: the November 6 manufacturing multi-agent paper uses a factory demonstrator and benchmarks, not a named operational production deployment.
Retail primary-source gap remains for Pick n Pay; Runnings is announced and not yet deployed. Walmart freight optimization lacks explicit learned-AI evidence.
Latin America, mainland China and non-English sources beyond the Japanese logistics pilot were not systematically covered.
UAMS Luma Agent Launchpad appears adjacent to existing research_f410523231a70da7a8774504, which is a Qventus initiative. To avoid an unsupported cross-vendor merge, the distinct named Mile Bluff deployment is the candidate from that source.
Sources with November 9 dates or future planned launches were not admitted as in-window deployments.
Ducos operational evidence supports Litmas use, not proof that newly launched Agentic Associate features are already adopted.
Sentara case-management primary release is October 20 and secondary reporting October 26, outside this incremental window; not counted as new.
No independent audit of claimed productivity or cash effects was found; throughput and task-time metrics are labelled source claims.
36 executed queries: 20 frontier-vendor searches and 16 regulator searches. Discovery is bounded, not exhaustive.
No qualifying new regulator-owned deployed/pilot AI workflow established. FCA November roundup recycles the October 19 speech https://www.fca.org.uk/news/speeches/building-future-finance-age-ai (date independently checked).
MAS FFI began operating November 3, but its PathFin.ai/BuildFin.ai/SafeFin.ai pillars are enablement/research/governance programmes, not evidence of a deployed regulator AI workflow: https://www.mas.gov.sg/news/media-releases/2026/future-of-finance-institute-commences-operations-starting-with-ai-in-finance .
HKMA November 2 Project SAGE is governance/exploration; primary direct retrieval timed out, retained only as coverage context: https://www.hkma.gov.hk/eng/news-and-media/press-releases/2026/11/20261102-4/ .
DFSA survey adoption aggregates and October 30 cyber-resilience paper do not identify a new regulator deployment. TPR historical scam-website work and HKIA aggregate case counts were not converted to new named deployments.
No fresh named OpenAI/Anthropic FS customer workflow established. Itaú October 16, Anthropic Barclays October 1, and Google Deutsche Bank August 18 are outside this window; future Anthropic November 10 event excluded.
Microsoft Bankwest November 4 is already owned by banking-refresh; omitted here to prevent duplication.
AWS Apexon November 2 article offers architecture guidance without a named bank deployment. Google November 4 PayPal reference remains a lead.
Verisk AI-SDLC November 5 is distinct from October 28 customer-facing insurance-data connectors; features of its five-agent engineering initiative are consolidated into one case.
Fulton has no baseline case in the supplied snapshot; no exact-ID merge. Verisk/Nammu21 baseline search found none.
No universal-coverage claim: English and Portuguese-led searches; limited local-language Asian, African and Middle Eastern coverage.
No newly deployed named pension-fund workflow verified. TPR October 29 guidance and Hong Kong IA November 2 aggregate incubation figures were passed to parent as regulatory context, not insurer deployments.
Manulife Advisor AI Assistant primary launch is October 14; Travelers Upload & Go, Verisk AI integrations, PolicyMe copilot and Insurify gateway releases are October 28; do not count later news repeats as fresh launches.
CNP Seguradora Ísis November 3 syndicated metrics repeat a full-text AWS case dated May 13, 2026; excluded from fresh candidates. AVALIA is separately named and its relationship to Ísis is unproven.
Gallagher Re November 5 report page was read but its 45 MB PDF failed web retrieval; no deployment candidates inferred from its aggregate funding figures.
Guardian go-live remains unconfirmed. SCOR initiative identity and primary transcript remain unresolved. Unimed metric definitions need reconciliation.
Primary GlobeNewswire AVALIA page returned cache miss; identifiable full syndicated vendor release used with explicit provenance.
BoundAI vendor launch claim lacks a named insurer deployment and exact November day; no case promoted.
Delegation attempted but agent thread limit prevented parallel regional research.
22/22 records assessed; 12 supported/corrected inclusion candidates, 3 partial held leads, 5 unsupported and 2 duplicate records. Assessment coverage is not verification coverage.
Ant/Zhima primary full text and explicit AI basis were not established; current scale and inclusion impact remain unverified.
Aladdin core ML-risk deployment and UBS NLP ESG-scoring pilot remain unconfirmed.
Exact BofA recruiting, Fidelity AI retirement, MetLife engagement-analytics pilot, SocGen unified risk platform and State Street AI report-generation claims remain unsupported.
Most sources are historical or undated; none is presented as newly announced within the incremental window.
No verified workflow-specific ROI metric was imported. Current deployments and precise dates remain unconfirmed where sources are historical/undated.
Nomura primary PDF full text was available but screenshot retrieval was blocked.
22/22 assigned records assessed. Nine legacy deployment updates, one separately identified Schwab retail initiative and one canonical Two Sigma enrichment are proposed; 13 legacy leads remain held. Assessment does not imply verification of every original claim.
Several historical deployments lack current-state confirmation; unknown publication and rollout dates remain null.
Ant International/NRI/Citadel Securities evidence must not silently transfer to Ant Group/Nomura Holdings/Citadel LLC.
BlackRock remains held for potential investment-signal overlap. Two Sigma NLP is recommended for catalog-level consolidation into canonical ML-Driven Investment Signals; this is not issuer-confirmed record identity.
Schwab advisor recommendations remain unsupported; actual retail Portfolio Insights is separate without a legacy-ID merge hint. UBS is narrowed to LAIA under exact original-citation continuity; original compliance automation and guarantees remain unsupported.
Morgan Stanley ML basis relies on indexed-primary brochure text while full-text SEC filing confirms operational use.
Direct source opens were performed in addition to the 39 search queries recorded below; inaccessible originals were replaced where primary evidence could be found.
No legacy impact or ROI label is reaffirmed without a supporting measurement.
All 21 exact claims assessed; assessed is not synonymous with verified.
22 primary-auditor and 22 bounded-checker search queries (44 total), plus direct opens/finds.
Eight source-backed corrections and one supported workflow proposed; 12 records remain leads including two duplicate-resolution cases.
Undated issuer pages retain null dates where unestablished.
Legacy redirects, source/claim mismatches, non-AI automation and research-only evidence retained.
No independent causal ROI verification and no conversion of capacity or payment-volume measures to savings.
Raymond James PDF full text opened by bounded checker; independent retry hit tool file-size cap.
All 21 assigned records assessed: 9 corrected merge candidates, 6 partial held leads and 6 unsupported held leads. A separate Global Sanctions Manager candidate is included without legacy overlap. Reviewed does not mean verified.
65 search queries executed across the lane and its three read-only research roles; source opens/finds excluded.
No newly deployed initiative in the 2026-10-29–2026-11-08 window is established by this lane.
Crédit Agricole regional-bank identity requires resolution before any merge into the S.A. entity.
Undated sources retain null publication dates and status dates; historical production is not asserted as current continuity.
Exact source-mismatched assertions remain held. Independent integrator adjudication is required.
Eight original claims remain unsupported: Bank of America ATM maintenance, Barclays market-risk AI, BlackRock talent analytics, Deutsche credit portfolio AI, the Fidelity entity/source mismatch, Goldman AI transaction monitoring, Nomura AI execution, and Societe Generale AI execution analytics.
BNP's specific bundled operational pilot and Wells Fargo's precise financial-crime system remain partial evidence holds. Nomura's non-AI-established algorithmic workflow is also retained only as a lead.
Barclays' original issuer source was blocked. Some original URLs were unavailable or mismatched; replacements are explicitly identified rather than assumed equivalent.
Historical deployment is distinguished from verified current operation. Persado renewal, current Progressive model implementation, and many exact launch dates are not established.
No universal institution coverage or realized financial ROI is asserted. Metrics are attributed usage/performance claims with explicit limitations.
Baseline institution/workflow titles were inspected. Exact original IDs are preserved for all proposed merges and held leads; unrelated named initiatives are not substituted for unsupported claims.
Fidelity Investments U.S. Virtual Assistant is a separate candidate with no original-ID overlap; it does not resolve the Freya/Fidelity International legacy mismatch.
47 search queries plus primary-source opens; bounded discovery, not universal institution or language coverage.
No verified new post-trade or private-equity internal deployment identified in this window. Tradeweb update concerns pricing, not deterministic AiEX automation.
Japan, India, Australia, Brazil and South Africa searches yielded mostly older material, surveys or AI-investment themes. They were not converted into deployments.
Amundi new benefit claim may belong to existing research_445ee19e5b4cea6fa91da879; held for identity adjudication.
Solovis full release obtained through syndicated Business Wire copy; original wire blocked and Yahoo direct open rate-limited.
GF Securities full text explicitly attributes GF Headlines but original issuer-hosted page not located. Jiufang retained as secondary-only lead.
Secondary roundups contain date drift: Allvue/RSM called new in November by an aggregator, while issuer says 7 May 2026; FINNY and Tandems releases and ProAllocate article date to 27 October. Excluded from new-window candidates.
No candidate demonstrates independently verified realized monetary ROI. Amundi reports analysis-time reduction only; Tradeweb reports a sample comparison without extractable numeric reduction.
Schwab MCP retail planning was reported on 30 October, but no independently verified in-window primary deployment update established in this bounded pass; not merged into Charley.
31 search queries plus primary page opens/clicks; broad regional searches do not constitute exhaustive country coverage.
Cashfree primary blog uses inconsistent relative publication dates; calendar date remains null. November 6 launch is supported by dated press coverage, with current availability confirmed by primary text.
No new treasury-specific or AML/KYC production case met explicit-AI and primary-evidence requirements in this bounded search.
Africa, Latin America and Japan yielded limited direct in-window evidence; absence here is not absence of deployment.
Natural is held as infrastructure availability without actual AI execution evidence; Payment Asia and Sumsub partnerships held pending live usage.
Chainalysis October 30 security notice, updated November 7, reports a service outage and restoration. This is contextual operational risk, not proof that existing Coinbase/Alterya AI fraud deployment stopped; no unsupported baseline downgrade made.
Baseline exact institution/workflow search found no matching records for three proposed additions; existing Klarna entries concern different workflows.
11 actual search queries plus primary-source opens/finds; bounded US gap review, not global coverage.
Eisen May 19 and founder July 23 sources are historical. No new in-window Eisen release was established.
Named customers corroborate Eisen platform use; neither customer lists nor BankProv testimonial prove which customers use each AI component.
Chimo evidence is self-description and partial-name testimonials; independent outcomes and technical details absent.
Texas October 30 is a genuine in-window AI announcement, but the source defers implementation detail.
North Carolina report cover gives July 2025 without a day; publicationDate is null, month recorded in statusBasis.
Utah 2026 agency presentation lists AI solutions alongside budget requests, without enough evidence to distinguish current phone-note deployment from plans.
No exact institution/workflow match found in baseline for proposed Eisen and North Carolina additions.

Search gaps addressed

The wealth/asset-management lane had not independently searched the large private-equity, private-credit and venture-capital cohort. — The final private-capital lane searched Blackstone, KKR, EQT, Carlyle, Apollo, Ares, Brookfield, Bain, Warburg Pincus, Advent, TPG, Partners Group and selected VC platforms. It produced 20 candidate records before final consolidation. KKR and Brookfield still have evidence gaps; global private-capital saturation is not claimed.
The initial payment/fintech sweep did not systematically cover business-finance platforms and digital-asset operators' internal financial workflows. — The final business-finance/digital-assets lane searched Ramp, Brex, BILL, Airwallex, Mercury, Navan, Melio, Coinbase and the named exchange/custody/analytics cohort. It added specific expense, AP, onboarding, service, fraud and risk-research candidates. Custody controls and several named operators remain unresolved rather than assumed absent.
Internal cyber defense, recruiting, audit, financial controls and named treasury adoption were thin across the initial sector lanes. — The final internal-functions lane searched those functions and supplied named cyber-defense, recruiting/career, audit, AP and treasury-adopter evidence. This closes the missing search pass, not every institution/function cell; ledger close, continuous control testing and specialist-asset-manager evidence remain thin.
Confiar's historical cybersecurity case was an unexamined cross-lane pointer in the emerging-market notes. — The internal-functions lane retrieved the named Confiar/Sofistic/Darktrace operating case. It is historical vendor-supported evidence, not proof of current 2026 scope or independent detection effectiveness.
Regional banking lanes intentionally omitted most trading, capital-markets and specialist investment workflows. — Dedicated capital-markets, wealth/asset-management and private-capital lanes searched those workflows. Their absence within a banking lane is not a global category absence, although trading-model and investment-return evidence remains selective.
Tax, civic administration, health and education were deliberately outside the financial-regulator lane. — Public/scientific and healthcare/life-sciences lanes examined those original-sector workflows for explicit FS transfer. They are not relabeled as existing FS deployments.

Unverified leads awaiting qualification (275)

These leads are separate from the catalog and do not establish a deployment.

Agricultural Bank of China Limited · AI credit-risk early-warning exploration
Institution and business context known, but deployed AI contribution to this workflow is insufficiently specific.
Lead sourceabchina.com · opens in a new tab
Bank of China (Hong Kong) Limited · BoC Pay+ intelligent consumption pilot
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebocg-life.bocgroup.com · opens in a new tab
Bank of China (Hong Kong) Limited · AI chat and hotline support
Clarify exact AI-operated functions and deployment boundaries before catalog inclusion.
Lead sourcebochk.com · opens in a new tab
Hang Seng Bank Limited · GenA.I. Sandbox++ AI-versus-AI pilot
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcehangseng.com · opens in a new tab
HDFC Bank Limited · HDFC Bank ONE omnichannel platform
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcev.hdfc.bank.in · opens in a new tab
KB Kookmin Bank · BICS corporate-loan assessment
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcekbfg.com · opens in a new tab
KB Kookmin Bank · AI sales-explanation compliance
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcekbinsure.co.kr · opens in a new tab
Kasikornbank Public Company Limited · Azure OpenAI employee-agent programme
Named customer story establishes adoption, but available retrieved content lacks sufficient task detail for an additional concrete use case.
Lead sourcemicrosoft.com · opens in a new tab
China Minsheng Banking Corp., Ltd. · LLM-assisted credit and customer-service workflows
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcewap.cmbc.com.cn · opens in a new tab
Shinhan Bank · Internal inquiry assistant with Microsoft
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebusinesskorea.co.kr · opens in a new tab
Woori Bank Vietnam Limited · FPT eID mobile onboarding
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefpteid.com · opens in a new tab
BitGo · Internal coding and productivity assistants
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcesec.gov · opens in a new tab
CipherTrace · Bank monitoring of high-risk crypto service providers
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefinance.yahoo.com · opens in a new tab
Kraken · Kai support chatbot evidence lead
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcereddit.com · opens in a new tab
TRM Labs · All-source investigative planning and evidence synthesis
Specific product but no named deploying organization in retrieved text; do not treat TRM's customer institutions as identified users.
Lead sourcetrmlabs.com · opens in a new tab
Lead sourcetrmlabs.com · opens in a new tab
The Depository Trust & Clearing Corporation · Corporate-action intelligent document-processing initiative
Named workflow but operational stage and implementation details remain unresolved; overall productivity claims must not attach to it.
Lead sourcefiles.dtcc.com · opens in a new tab
The Depository Trust & Clearing Corporation · AI data extraction and enrichment exploration
No deployment-specific stage established.
Lead sourcefiles.dtcc.com · opens in a new tab
Euronext N.V. · Scila surveillance offering to Euronext Athens members
Generic distributed capability lacks a named deployment; not evidence Euronext itself uses the AI modules.
Lead sourceathens.euronext.com · opens in a new tab
Euronext N.V. · Borsa Italiana AI surveillance prototyping
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceeuronext.com · opens in a new tab
Intercontinental Exchange, Inc. · Fixed-income data access through AI platforms
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceir.theice.com · opens in a new tab
JPMorgan Chase & Co. · AI synthesis and client-material preparation in investment banking
Real investment-banking workflow, but tool identity is unspecified. Preserve as pending workflow variant; neither count as a distinct initiative nor assert a merge into LLM Suite.
Lead sourcekitco.com · opens in a new tab
Abu Dhabi Commercial Bank PJSC · Enterprise AI transformation programme
Do not create 150 records or assign all programme cases production.
Lead sourceadcb.com · opens in a new tab
Grupo Financiero Banorte S.A.B. de C.V. · AI hyperpersonalization and predictive services partnership
Broad partnership scope; do not infer production personalization from infrastructure and model-development claims.
Lead sourcebanorte.com · opens in a new tab
BBVA México S.A. · ARIA anti-extortion-fraud initiative
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebbva.com · opens in a new tab
BBVA México S.A. · Blue telephone banking assistant in Mexico
Retrieve original Q1 2026 Mexican disclosure and specific deployment article.
Lead sourceexpansion.mx · opens in a new tab
BBVA México S.A. · AI card-fraud detection in Mexico
No full article examined; metric denominator and deployment details need confirmation.
Lead sourceitmastersmag.com · opens in a new tab
Banco de Chile · Microsoft 365 Copilot adoption agreement
Specific deployed employee workflow and rollout stage not established in examined excerpt.
Lead sourcesitiospublicos.bancochile.cl · opens in a new tab
Ecobank Transnational Incorporated · Rafiki chatbot for EcobankPay merchant onboarding
Workflow is clear, but AI evidence in primary page is missing; do not equate every chatbot with AI.
Lead sourceecobank.com · opens in a new tab
First Abu Dhabi Bank PJSC · AI-supported threat intelligence and cyber defense
Need full case to establish concrete AI mechanism and stage.
Lead sourcebankfab.com · opens in a new tab
First Abu Dhabi Bank PJSC · Agentic trade-document validation
Deployment stage unresolved; full case required.
Lead sourcebankfab.com · opens in a new tab
Banco Internacional del Perú S.A.A. · Claude enterprise and agentic AI agreement
Access agreement is not a named workflow deployment.
Lead sourceinterbank.pe · opens in a new tab
Bank Leumi Le-Israel B.M. · Generative AI hub
Hub and strategic intent, not a concrete deployed workflow.
Lead sourcecalcalistech.com · opens in a new tab
The Saudi National Bank · SymphonyAI sanctions screening and transaction monitoring
Vendor company name alone does not prove deployed AI functionality.
Lead sourcealahli.com · opens in a new tab
Barclays · Colleague AI productivity / Copilot rollout
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcehome.barclays · opens in a new tab
CaixaBank · AI4CYBER anomaly-detection project
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceopenmobile.caixabank.es · opens in a new tab
Crédit Agricole / LCL · ARIA adviser assistance
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcecredit-agricole.com · opens in a new tab
Danske Bank · ML payment fraud scoring
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceassets.teradata.com · opens in a new tab
Deutsche Bank / Postbank · Postbank customer chatbot Paula
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcesec.gov · opens in a new tab
Eurobank · Agentic AI factory for core banking operations
Accessible announcement identifies platform and future core-operations automation, but no specific operating workflow; retained as lead.
Lead sourceeurobank.gr · opens in a new tab
OTP Group · Hungarian language-model platform for bank services
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceotpgroup.info · opens in a new tab
Raiffeisen Bank International · Network-bank voicebots and CRM-connected chatbot
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcerbinternational.com · opens in a new tab
UniCredit Bank Serbia · AI customer churn and downgrade modeling
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceunicreditbank.rs · opens in a new tab
ANZ Bank · GoBiz external-data lending orchestration
Data orchestration and automation established; specific AI inference within this workflow not established.
Lead sourcefeedzai.com · opens in a new tab
BigPay · Cross-channel mule detection rules
Specific response documented as analyst-authored rules and simulations; cannot attribute stated mule reduction to AI merely from vendor identity.
Lead sourcefeedzai.com · opens in a new tab
Currencycloud · Actor Intelligence investigation analytics
Real named operating workflow, but source does not clearly establish which AI inference mechanism was deployed; issuer press link returned loading shell.
Lead sourcelucinity.com · opens in a new tab
Fiserv · Governed enterprise automation program
Broad program evidence is insufficient to connect a specific financial-services workflow to the AI component.
Lead sourceuipath.com · opens in a new tab
HSBC Holdings plc · Insurance transaction monitoring with ARIC machine learning
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefeaturespace.com · opens in a new tab
Hampshire Trust Bank · Transaction monitoring and customer/payment screening
Named deployment confirmed but source focuses on rules/data; AI-specific workflow attribution needs corroboration.
Lead sourcecomplyadvantage.com · opens in a new tab
Mambu · Agentic platform product capability
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcedocs.mambu.com · opens in a new tab
NatWest Group · Adaptive machine-learning scam detection
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefeaturespace.com · opens in a new tab
Oracle Financial Services · Retail-banking AI agent platform launch
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceoracle.com · opens in a new tab
Raiffeisen Bank International · CRISP reconciliation platform upgrade
Named reconciliation deployment confirmed, but BRx Match generic AI capability does not prove client deployed its AI module.
Lead sourcebroadridge.com · opens in a new tab
Standard Chartered PLC · Retail onboarding external-data orchestration
General bank AI use is mentioned, but the named workflow itself is API/data orchestration; insufficient to classify as AI deployment.
Lead sourcefeedzai.com · opens in a new tab
Standard Chartered PLC · Connected customer intelligence for financial crime investigations
Reviewed: Identity review: a source supports this named workflow, but the broad baseline record may describe the same initiative. The baseline source could not be matched in this research pass. Preserve the evidence as an uncounted workflow variant; do not invent distinction or assert a merge.
Lead sourcequantexa.com · opens in a new tab
Lead sourcequantexa.com · opens in a new tab
STET · AI transaction fraud partnership
Named partnership but operational deployment detail is video-only and was not recovered.
Lead sourcefeedzai.com · opens in a new tab
Temenos · Unnamed Tier 1 bank sanctions-screening AI agent
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceglobenewswire.com · opens in a new tab
Deutsche Bank · Vault Core modernization with AI integration potential
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcethoughtmachine.net · opens in a new tab
Equifax Inc. · Equifax alternative-data credit scoring
Reviewed: Offering availability is established, but this source does not demonstrate a named client deployment or independently establish the provider operating this workflow. Preserve product evidence without counting it as observed institutional adoption.
Lead sourceinvestor.equifax.com · opens in a new tab
Regions Financial Corporation · Recruiting personalization and engagement automation
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcephenom.com · opens in a new tab
GSK · GSK historical quality-investigation analysis
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcegsk.com · opens in a new tab
Mayo Clinic · Mayo Abridge enterprise expansion
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcehealthcaredive.com · opens in a new tab
Lead sourceabridge.com · opens in a new tab
Mayo Clinic · Mayo patient-safety incident classification and search
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcepmc.ncbi.nlm.nih.gov · opens in a new tab
Mount Sinai Health System · Mount Sinai planned surgical-supply capture
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceprnewswire.com · opens in a new tab
F. Hoffmann-La Roche · Roche pharmaceutical visual-inspection validation
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceema.europa.eu · opens in a new tab
Airbus SE · Airbus experiments with a Standard Operating Instructions assistant
Specific experiment is documented, but operating deployment and subsequent production status are not established.
Lead sourceairbus.com · opens in a new tab
BP plc · bp announces AIP decision support over its production digital twin
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcepalantir2020ipo.q4web.com · opens in a new tab
TotalEnergies SE · RAID centrally forecasts rotating-machine failures
Operational workflow exists, but AI/ML identity is not demonstrated sufficiently in the specific source.
Lead sourcecstjf-pau.totalenergies.fr · opens in a new tab
Baker McKenzie · Legora rollout with Applied AI workflow co-development
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcebakermckenzie.com · opens in a new tab
Bowmans · Legora African-law-firm rollout
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcelinkedin.com · opens in a new tab
Deloitte · Relativity/Brainspace document prioritisation in corruption investigation
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceai-hungary.com · opens in a new tab
Deloitte · Omnia GenAI documentation, statement review and memo capabilities
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcedeloitte.com · opens in a new tab
DLA Piper · Harvey global rollout after multi-practice pilot
Adoption is evidenced but concrete used workflow is not specified in the fetched article.
Lead sourceharvey.ai · opens in a new tab
EY · Tax Agent Factory and Tax Labs launch
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceey.com · opens in a new tab
Harneys · Harvey cross-jurisdiction precedent review and drafting rollout
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceharneys.com · opens in a new tab
Pinheiro Neto Advogados · Harvey customer listing
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceharvey.ai · opens in a new tab
PwC Belgium · Belgian-legislation Tax AI Assistant adoption
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcepwc.be · opens in a new tab
Wotton + Kearney · Lexis+ AI case-law research adoption
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcelexisnexis.com · opens in a new tab
Dai-ichi Life Holdings, Inc. · Dai-ichi Life sales-representative assistance
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcedaiichilife-group.com · opens in a new tab
FWD Group Holdings Limited · FWD bancassurance lapse winback targeting
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefwd.com · opens in a new tab
Manulife Financial Corporation · Manulife sales support in Singapore and Japan
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcemanulife.com · opens in a new tab
Nippon Life Insurance Company · Nippon Life proposed administrative question answering
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcelinks.sgx.com · opens in a new tab
Prudential plc · Prudential Malaysia GenAI medical-claim adjudication
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceprudentialplc.com · opens in a new tab
Prudential Financial, Inc. · Prudential Financial AI-assisted annuity pricing models
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcepublicnow.com · opens in a new tab
Prudential Financial, Inc. · Prudential Financial underwriting and claims acceleration
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcepublicnow.com · opens in a new tab
The Company for Cooperative Insurance (Tawuniya) · Tawuniya health preauthorization approval model
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcetawuniya.com · opens in a new tab
Citigroup Inc. · Citi Smart Match receivables application
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcecitibank.com · opens in a new tab
First National Bank of Omaha · Pindrop voice fraud and authentication
Specific AI/ML mechanism not established in reviewed customer-page passages; product functionality alone is not enough to assign AI.
Lead sourcepindrop.com · opens in a new tab
Fulton Bank · Contact-center oversight intelligence
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcelinkedin.com · opens in a new tab
Huntington Bancshares Incorporated · ML fraud-protection workflow details
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourced1io3yog0oux5.cloudfront.net · opens in a new tab
JPMorgan Chase & Co. · Persado marketing language optimization
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcepersado.com · opens in a new tab
JPMorgan Chase & Co. · AI threat-modeling co-pilot
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcejpmorgan.com · opens in a new tab
ORNL Federal Credit Union · AI underwriting association requires verification
Source associates an executive with historic adoption without clearly stating deployment at ORNL itself.
Lead sourcezest.ai · opens in a new tab
Royal Bank of Canada · ATOM credit-capacity inference
Identifiable research direction, but production credit decisioning not established.
Lead sourcerbcborealis.com · opens in a new tab
Wells Fargo & Company · AI-assisted software engineering
Reviewed: Identity review: source confirms an unnamed workflow but does not establish separate initiative identity from the broad baseline. Preserve as a workflow variant; do not merge or count as net-new without identity evidence.
Lead sourcestories.wf.com · opens in a new tab
Wells Fargo & Company · AI underwriting-document creation
Reviewed: Identity review: source confirms an unnamed workflow but does not establish separate initiative identity from the broad baseline. Preserve as a workflow variant; do not merge or count as net-new without identity evidence.
Lead sourcestories.wf.com · opens in a new tab
Fiserv, Inc. · AI-assisted client service and issue anticipation
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceinvestors.fiserv.com · opens in a new tab
Fiserv, Inc. · AI-assisted delivery, migration and testing
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceinvestors.fiserv.com · opens in a new tab
Mastercard Incorporated · Graph and generative risk scoring at authorization
Reviewed: Offering availability is established, but this source does not demonstrate a named client deployment or independently establish the provider operating this workflow. Preserve product evidence without counting it as observed institutional adoption.
Lead sourcenewsroom.mastercard.com · opens in a new tab
Lead sourcemastercard.com · opens in a new tab
Mercado Pago · Real-time payment risk scoring with review escalation
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcemercadopago.com.br · opens in a new tab
National Payments Corporation of India · Payment grievance and mandate support with a domain model
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcenpci.org.in · opens in a new tab
Lead sourcenpci.org.in · opens in a new tab
PayPal Holdings, Inc. · Payment acceptance from AI shopping assistants
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcedeveloper.paypal.com · opens in a new tab
PayPal Holdings, Inc. · Personalized merchant recommendations in receipts
Specific announced product; individual rollout status not resolved.
Lead sourcenewsroom.paypal-corp.com · opens in a new tab
Remitly Global, Inc. · Machine-learning pricing across remittance corridors
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcesec.gov · opens in a new tab
Visa Inc. · Enumeration-attack scoring for card-not-present authorizations
Full announcement; later operational availability not established in retrieved text.
Lead sourceinvestor.visa.com · opens in a new tab
Worldpay · AI selection of 3DS versus direct authorization
Reviewed: Offering availability is established, but this source does not demonstrate a named client deployment or independently establish the provider operating this workflow. Preserve product evidence without counting it as observed institutional adoption.
Lead sourceworldpay.com · opens in a new tab
Brookfield Business Partners L.P. · AI support for investment sourcing and diligence — opportunity statement
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebbuc.brookfield.com · opens in a new tab
KKR & Co. Inc. · KARA and enterprise GenAI productivity
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceir.kkr.com · opens in a new tab
American International Group, Inc. · Claims by AIG Assist
Reviewed: Identity review: a source supports this named workflow, but the broad baseline record may describe the same initiative. The baseline source could not be matched in this research pass. Preserve the evidence as an uncounted workflow variant; do not invent distinction or assert a merge.
Lead sourceaig.com · opens in a new tab
American International Group, Inc. · Underwriting by AIG Assist
Reviewed: Identity review: a source supports this named workflow, but the broad baseline record may describe the same initiative. The baseline source could not be matched in this research pass. Preserve the evidence as an uncounted workflow variant; do not invent distinction or assert a merge.
Lead sourceaig.com · opens in a new tab
Lead sourcesec.gov · opens in a new tab
Allstate Corporation · Allstate AI drafting of claimant emails
Only indexed reporting obtained in this pass; obtain full original reporting before inclusion.
Lead sourcenews.bloomberglaw.com · opens in a new tab
Chubb Limited · Chubb claim-severity and adjuster-action models
Require workflow-specific rollout evidence; capability architecture alone is insufficient.
Lead sourcecoverager.com · opens in a new tab
Chubb Limited · Chubb commercial intelligent intake and pre-underwriting
Concrete workflow, but source does not isolate which production steps are AI versus rules-based automation.
Lead sourcecoverager.com · opens in a new tab
Liberty Mutual Insurance Group · Liberty Mutual AI-assisted recruiting
Specific function but disclosure uses conditional language; needs a named actual hiring workflow or system.
Lead sourcelibertymutualgroup.com · opens in a new tab
Sompo Holdings, Inc. · Sompo Japan AIP underwriting knowledge support
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceyoutube.com · opens in a new tab
Swiss Re Ltd · Swiss Re AI contract-wording review
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceswissre.com · opens in a new tab
University of Cambridge · Smart-Review grant-funding contract comparison
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceai.cam.ac.uk · opens in a new tab
Lead sourceai.cam.ac.uk · opens in a new tab
Department of Public Service and Administration · Batho Pele public-servant policy knowledge assistant
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcesanews.gov.za · opens in a new tab
Income Tax Department, India · AI-assisted faceless tax-assessment case selection
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceincometaxindia.gov.in · opens in a new tab
Australian Prudential Regulation Authority · Copilot departmental productivity enablement
Primary evidence of adoption, but detailed deployed workflows are not disclosed.
Lead sourceapra.gov.au · opens in a new tab
Banco Central do Brasil · ADAM supervisory credit-file selection
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebcb.gov.br · opens in a new tab
Banco Central do Brasil · NLP document-processing supervisory tool programme
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebcb.gov.br · opens in a new tab
Bank for International Settlements · Ellipse integrated supervisory analytics prototype
Prototype and workflow verified, but specific AI mechanism needs full-text project-report verification.
Lead sourcedocs.bis.org · opens in a new tab
Lead sourcebis.org · opens in a new tab
Central Bank of the United Arab Emirates · SupTech risk-monitoring and AML analytics programme
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcecentralbank.ae · opens in a new tab
Reserve Bank Innovation Hub · MuleHunter bank mule-account detection
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcedocs.rbihub.in · opens in a new tab
Securities and Exchange Board of India · Sudarshan and Radar market-abuse surveillance tools
Primary speech/report and full secondary text needed; retain combined lead until separate tool scopes are verified.
Lead sourcemoneycontrol.com · opens in a new tab
Air France · Air France customer-feedback analysis
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcecorporate.airfrance.com · opens in a new tab
Carrefour S.A. · Carrefour Poland omnichannel service assistant
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceusekoda.com · opens in a new tab
Expedia Group, Inc. · Expedia Romie travel-planning assistant
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcephocuswire.com · opens in a new tab
H & M Hennes & Mauritz AB · H&M demand forecasting and replenishment pilot
Workflow is real, but accessible full text establishes an algorithm, not explicitly AI/ML for this particular system. Primary AI corroboration remained indexed only.
Lead sourceideo.com · opens in a new tab
Lead sourcehmgroup.com · opens in a new tab
Industria de Diseño Textil, S.A. · Inditex AI comparable-product demand forecasting
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcestatic.inditex.com · opens in a new tab
MercadoLibre, Inc. · Mercado Libre Mexico vehicle-marketplace transaction agents
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcenews.mercadolibre.com · opens in a new tab
Orange S.A. · Orange Sosh speech-to-speech service assistant
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceorange.com · opens in a new tab
T-Mobile US, Inc. · T-Mobile intent-driven customer-experience platform
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcet-mobile.com · opens in a new tab
Airtable · Hyperagent browser access and recorded web-action sessions
Keep as lead: operating platform and traffic are established, but no concrete named end-user workflow is documented.
Lead sourcebrowserbase.com · opens in a new tab
Vercel · Prism web intelligence and churn-signal collection
Keep as lead: named workflow is real, but source does not establish an AI implementation; browser automation alone is not AI.
Lead sourcebrowserbase.com · opens in a new tab
Amundi S.A. · Fund-document compliance consistency checker
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcecioreview.com · opens in a new tab
APG Group N.V. · APG self-service investment analytics preparation
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceapg.nl · opens in a new tab
BlackRock, Inc. · AIM combines investment signals into portfolio-aware alpha models
Reviewed: Identity review: the researched account is supported but the broad legacy initiative may encompass it. Source-aware review could not establish a distinct initiative or exact match. Preserve as a qualified workflow variant pending identity resolution.
Lead sourceblackrock.com · opens in a new tab
BlackRock, Inc. · Systematic earnings-call NLP and thematic baskets
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceblackrock.com · opens in a new tab
Fidelity Investments · Saifr marketing compliance review platform
Actual product launch is clear, but no named operational client or internal rollout established by this source.
Lead sourcefintechfutures.com · opens in a new tab
Morgan Stanley · Next Best Action advisor recommendations
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebusiness.columbia.edu · opens in a new tab
Nomura Holdings, Inc. · Nomura Deep Research for investment advice and analysis
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcenomuraholdings.com · opens in a new tab
Royal Bank of Canada · AI-Powered Insights identifies clients with changing financial needs
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcerbcwealthmanagement.com · opens in a new tab
T. Rowe Price Group, Inc. · Chat TRP and Claude support investment research and tool development
Reviewed: Identity review: the researched account is supported but the broad legacy initiative may encompass it. Source-aware review could not establish a distinct initiative or exact match. Preserve as a qualified workflow variant pending identity resolution.
Lead sourcetroweprice.com · opens in a new tab
Lead sourceclaude.com · opens in a new tab
T. Rowe Price Group, Inc. · Investment Data Insights applies NLP and predictive research
Reviewed: Identity review: the researched account is supported but the broad legacy initiative may encompass it. Source-aware review could not establish a distinct initiative or exact match. Preserve as a qualified workflow variant pending identity resolution.
Lead sourcetroweprice.com · opens in a new tab
Two Sigma Investments, LP · AI-assisted feature engineering and research workflows
Reviewed: Identity review: the researched account is supported but the broad legacy initiative may encompass it. Source-aware review could not establish a distinct initiative or exact match. Preserve as a qualified workflow variant pending identity resolution.
Lead sourcetwosigma.com · opens in a new tab
Abu Dhabi Commercial Bank · Board Observer: leadership decision support
AI basis not sufficiently recorded. The issuer explicitly says deployed.
Lead sourceadcb.com · opens in a new tab
Abu Dhabi Commercial Bank · Enterprise Knowledge: employee policy and regulatory-document retrieval
AI basis not sufficiently recorded. Named tool and staff workflow explicitly reported.
Lead sourceadcb.com · opens in a new tab
Attijariwafa bank / Simple (Morocco) · Conversational account support and customer-confirmed banking transactions
Researcher decision: lead. Promising recent local-language evidence, but primary-source inclusion gate unmet.
Lead sourcemedias24.com · opens in a new tab
Lead sourcesimple.ma · opens in a new tab
First Abu Dhabi Bank PJSC · Compliance assurance information review and risk-based assessment support
Potential baseline initiative overlap lacks unique identity proof; retained for adjudication. Operating function, task and oversight directly supported.
Lead sourcebankfab.com · opens in a new tab
Lead sourcebankfab.com · opens in a new tab
First Abu Dhabi Bank PJSC · Cyber-advisory triage, exposure validation and owner-mapped response preparation
Potential baseline initiative overlap lacks unique identity proof; retained for adjudication. In-use workflow visually verified; autonomous remediation is not inferred.
Lead sourcebankfab.com · opens in a new tab
Lead sourcebankfab.com · opens in a new tab
Hanwha Life Insurance · Employee Financial Data Agent for querying, analysis and report preparation/distribution
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcefintechcenter.or.kr · opens in a new tab
Samsung Fire & Marine Insurance · Machine-learning long-term health underwriting using disclosed conditions and claims history
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcesamsungfire.co.kr · opens in a new tab
DNB · Agentic customer-data checks and KYC
Researcher decision: lead. Named workflow but source wording does not establish KYC-specific production stage precisely enough.
Lead sourcenews.cision.com · opens in a new tab
Amazon AGI Lab · Training and evaluating UI agents against backend-state verifiers
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceamazon.science · opens in a new tab
Anthropic · Internal purchase-request review agent
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceanthropic.com · opens in a new tab
Anthropic · Agent-assisted source-control migration and component health monitoring
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceanthropic.com · opens in a new tab
Google DeepMind / GovTech Singapore · Privacy-preserving local scam-conversation detection
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcedeepmind.google · opens in a new tab
Meta · Agent-session-aware data access observation and enforcement
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceengineering.fb.com · opens in a new tab
Georgia Farm Bureau · Personal-lines actuarial territory modeling and rating justification
AI basis not sufficiently recorded. Specific older omission: small US regional actuarial team and named operating workflow.
Lead sourceakur8.com · opens in a new tab
Branch Insurance Exchange · Regulatory filing extraction for actuarial pricing research
AI basis not sufficiently recorded. Distinct US insurance pricing-research deployment, absent baseline.
Lead sourceakur8.com · opens in a new tab
EIOPA survey of 347 insurers across 25 countries · GenAI deployment maturity and operating-risk controls
Context only: EIOPA survey is cross-sector maturity and risk evidence, not a named operating workflow.
Lead sourceeiopa.europa.eu · opens in a new tab
Aon / SAGE / AgriCap / RALIC · Field-level precision crop underwriting
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceaon.mediaroom.com · opens in a new tab
Swiss Re Corporate Solutions · ClaimsGenAI fraud/recovery opportunity triage
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceswissre.com · opens in a new tab
Lead sourceswissre.com · opens in a new tab
WTW ResQ Labs · Reserving automation script development and debugging
Researcher decision: lead. Useful exact reserving workflow lead; customer-level operating evidence still missing.
Lead sourcewtwco.com · opens in a new tab
Guidewire (platform supplier) · Agentic underwriting, claims and actuarial pricing platform
Researcher decision: lead; AI basis not sufficiently recorded. Fresh primary launch but cannot be promoted to named insurer operations.
Lead sourceir.guidewire.com · opens in a new tab
Banco Bci · Legal and regulatory query assistant
Researcher decision: lead; AI basis not sufficiently recorded. Specific workflow currently rests on indexed interview text, so do not promote as fully verified.
Lead sourcedf.cl · opens in a new tab
Lead sourcecdn3.bci.cl · opens in a new tab
Banco Industrial · ABI voice banking transactions
AI basis not sufficiently recorded. Distinct deployed workflow supported by primary full text.
Lead sourceblog.corporacionbi.com · opens in a new tab
Lead sourceembed-rech-01.dialog.cm · opens in a new tab
Banco Industrial · CISBOT internal banking-query routing and answers
AI basis not sufficiently recorded. Distinct deployed workflow supported by primary full text.
Lead sourceembed-rech-01.dialog.cm · opens in a new tab
Banco do Nordeste do Brasil · LuzIA credit-policy knowledge assistance
AI basis not sufficiently recorded. Distinct deployed workflow supported by primary full text.
Lead sourceinvestidor10.com.br · opens in a new tab
Diversified Members Credit Union · ML origination-fraud and anomaly detection
Researcher decision: lead. Capture the concrete selected solution while withholding production status.
Lead sourcefintech-pulse.com · opens in a new tab
Lead sourcescienaptic.ai · opens in a new tab
FORUM Credit Union · AI summarization of commercial property appraisals
Researcher decision: lead. Strong named non-vendor interview, but technical AI basis needs confirmation.
Lead sourcecreditunions.com · opens in a new tab
Huntington National Bank · 1866 Studio small-business launch and loan-readiness agents
Researcher decision: lead. Concrete named SME workflow and model family, but signup is not deployment evidence.
Lead sourceir.huntington.com · opens in a new tab
Regions Bank · AI-assisted lockbox research, archiving and transaction completion
Researcher decision: lead. Useful treasury gap lead, held below inclusion threshold because technical AI basis is missing.
Lead sourcebai.org · opens in a new tab
HSBC Holdings plc · Extract and check documentary-trade information; route uncertain results to specialists
Possible overlap with existing HSBC AI Trade Finance Processing (cmmpo6ij2004nvwfg9v7sh9j1); same documentary-checking mechanism but unique product identity not established. Retain lead pending exact identity review.
Lead sourcebusiness.hsbc.com · opens in a new tab
Lead sourcedisruptionbanking.com · opens in a new tab
J.P. Morgan Payments / Prysmian · Cash Flow Intelligence forecasting and reconciliation
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcejpmorgan.com · opens in a new tab
Lloyds Bank · ML/NLP extraction and examination of trade-finance documents
Researcher decision: lead. Named-bank primary evidence, retained as follow-up until actual implementation is confirmed.
Lead sourcelloydsbankinggroup.com · opens in a new tab
Ripple Treasury · GSmart orchestration across cash forecasting, reconciliation, liquidity and payments
Researcher decision: lead. Recent primary announcement worth following, insufficient adopter/workflow evidence for inclusion.
Lead sourcetreasury.ripple.com · opens in a new tab
RockWallet / MNEE App · Crypto/fiat on-ramp fraud controls and bot attack detection
Researcher decision: lead. Real digital-asset operations, but missing explicit AI attribution prevents inclusion.
Lead sourcesardine.ai · opens in a new tab
Capita plc / Civil Service Pension Scheme · Promised AI-enabled pension administration and recovery automation
Context only: official failed-delivery and operational-control countercase; does not establish a deployed AI workflow or isolate an AI model as cause.
Lead sourcehansard.parliament.uk · opens in a new tab
Lead sourcecomputerweekly.com · opens in a new tab
Autorité de contrôle prudentiel et de résolution · LUCIA AML/CFT on-site examination and transaction selection
AI basis not sufficiently recorded. Named authority, concrete inspection workflow, explicit operational wording and primary full text.
Lead sourcebanque-france.fr · opens in a new tab
Lead sourceacpr.banque-france.fr · opens in a new tab
BaFin, Germany · ALMA ML-enhanced market-abuse detection
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcebafin.de · opens in a new tab
Blackstone Inc. · Microsoft 365 behavioural threat detection
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcevectra.ai · opens in a new tab
Lead sourcecdn.prod.website-files.com · opens in a new tab
PT Bank Hibank Indonesia · AI threat detection and managed response
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcevectra.ai · opens in a new tab
Lead sourcecdn.prod.website-files.com · opens in a new tab
The Charles Schwab Corporation · Client-facing portfolio performance explanations
Retail 2026 product differs in audience from legacy AI Portfolio Insights for Advisors (cmmpo6iy10057vwfgip81dtyj), but legacy identity is weak. Preserve as lead rather than silently correcting or duplicating without identity proof.
Lead sourcepressroom.aboutschwab.com · opens in a new tab
SS&C Technologies Holdings, Inc. · AI-assisted NAV and allocation review in GlobeOp
Product-availability lead: no named customer operating deployment or distinct SS&C internal operating workflow established.
Lead sourceinvestor.ssctech.com · opens in a new tab
The Pensions Regulator · Oversight of AI used by trustees and pension providers
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcethepensionsregulator.gov.uk · opens in a new tab
Anthropic Frontier Red Team; fixes acknowledged by Google and Apple · AI-assisted vulnerability discovery, verification, rating and coordinated disclosure
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceanthropic.com · opens in a new tab
Lead sourcechromereleases.googleblog.com · opens in a new tab
Ofsted (UK) · Inspection-evidence-to-report drafting assistant discontinued after beta
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcegov.uk · opens in a new tab
Ofsted (UK) · Survey-theme and safeguarding-flag analysis evaluated away from live decisions
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcegov.uk · opens in a new tab
Anthropic / Theorem · LLM-written machine-checked proofs for compiled fuzz-test entry points
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceanthropic.com · opens in a new tab
Lead sourcegithub.com · opens in a new tab
ING Group · Katana dealer bond-quote recommendations
AI basis not sufficiently recorded. Older missed primary evidence of learned pricing support; preserve historical pilot stage.
Lead sourceing.com · opens in a new tab
JPMorgan Chase & Co. · Quantum deep-learning and reinforcement-learning hedging research
Researcher decision: lead. Research evidence only; retain as a derivatives-risk lead rather than deployed capability.
Lead sourcejpmorganchase.com · opens in a new tab
Societe Generale S.A. · SG Markets My Cases post-trade request triage
Potential baseline initiative overlap lacks unique identity proof; retained for adjudication. Specific NLP role in post-trade operations.
Lead sourcesocietegenerale.com · opens in a new tab
Societe Generale S.A. · Trading Assistant derivatives RFQ parsing
Potential baseline initiative overlap lacks unique identity proof; retained for adjudication. Evidence enrichment or distinct workflow, subject to registry deduplication.
Lead sourcesocietegenerale.com · opens in a new tab
State Street Corporation · ML predictive reconciliation matching and break resolution
Possible overlap with existing AI-Powered Custody Operations (cmmpo6kjq0077vwfg51xl1e8n), whose description already includes ML reconciliation and exception handling. Retain lead pending unique initiative identity.
Lead sourcestatestreet.com · opens in a new tab
Ashurst · Blinded quality comparison and staff experiments for legal first drafts and case summaries
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceashurst.com · opens in a new tab
Bracewell LLP; Vorys, Sater, Seymour & Pease LLP; Cole, Scott & Kissane · Early-access agent-assisted litigation evidence investigation
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceeverlaw.com · opens in a new tab
Evidently / unnamed health systems · Longitudinal record context delivered to referral queues and other agents
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourceevidently.com · opens in a new tab
Legatics / unnamed law-firm customers · AI-assisted creation/update of matters and deal checklists through permissioned MCP writes
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcelegatics.com · opens in a new tab
Memorial Health System (the mhsystem.org organization linked by the source) · Appeal evidence assembly, multi-level outcome tracking and root-cause feedback
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcecofactorai.com · opens in a new tab
ING · Agentic support for complex Dutch mortgage approvals
Hold: original transcript or call passage not independently checked in the bounded primary-source search.
Lead sourceroic.ai · opens in a new tab
Rex Microfinance Bank Limited · RexAtlas personalized banking guidance
Hold: launch timing and currently deployed Atlas scope require reliable primary corroboration.
Lead sourcecryptocompass.com · opens in a new tab
Lead sourcerexafrica.com · opens in a new tab
Santander Portugal · Noxus knowledge assistant for product and compliance queries
Hold: refresh-window news found, but original statement and material novelty versus older vendor deployment claims not established.
Lead sourcejornaleconomico.sapo.pt · opens in a new tab
Pick n Pay · OrderGrid store-fulfilment pilot and rollout need primary corroboration
Hold: full secondary article read, but primary technical and deployment evidence was not located.
Lead sourcetechcentral.co.za · opens in a new tab
Runnings · Unified forecasting and replenishment rollout not yet live
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcerelexsolutions.com · opens in a new tab
Lead sourcefinance.yahoo.com · opens in a new tab
CERSI-AI · Federated clinical-AI performance monitoring plan
Cross-industry transfer needs an actual pilot or deployed workflow and explicit application, rationale and transfer limits.
Lead sourcecersi-ai.org · opens in a new tab
Walmart Inc. · First-mile freight optimization requires AI-basis adjudication
Hold: deployed optimization is real, but affiliation with Applied AI does not demonstrate an AI method; do not count as verified AI.
Lead sourcetech.walmart.com · opens in a new tab
Financial Conduct Authority · Planned quantum-machine-learning financial-crime experiments
Future experiment: launch is planned for Q1 2027, with participation still open. No current pilot or deployment.
Lead sourcefca.org.uk · opens in a new tab
Nammu21 · AI credit-agreement structuring for programmable syndicated-loan lifecycle
Hold for deployment eligibility: source models a loan proof of concept but identifies no live bank participant, borrower or operational pilot.
Lead sourcecloud.google.com · opens in a new tab
PayPal Holdings, Inc. · Agentic Data Cloud customer reference requiring workflow confirmation
Customer name alone does not establish a specific deployed AI use case or a new material update.
Lead sourcecloud.google.com · opens in a new tab
PolicyMe · PolicyMe complex-case underwriting copilot remains in beta
Held as a status-correction lead, not a new launch; no matching PolicyMe baseline case found.
Lead sourceadvisor.ca · opens in a new tab
Lead sourcefinance.yahoo.com · opens in a new tab
SCOR SE · SCOR augmented underwriting pilot retains human offer authorization
Material pilot-status account, held pending primary corroboration and exact relationship to existing SCOR terrorism record; no feature-splitting.
Lead sourceinsuranceasianews.com · opens in a new tab
Seguros Unimed · ANA assists Nascer Seguro maternity-program enrollment
Hold for primary evidence and metric reconciliation; retained because November disclosure adds explicit observation period but pre-window coverage exists.
Lead sourceterra.com.br · opens in a new tab
Lead sourcesaudedigitalnews.com.br · opens in a new tab
UBS Group AG · Exploratory AI for sustainable-finance analysis
UBS's sustainability report supports developing and exploring AI for ESG questionnaires and sustainability assessments, not the exact NLP scoring pilot. Original award article's ESG data platform and separate smart-building AI do not establish the asserted combination.
Lead sourcesec.gov · opens in a new tab
BlackRock, Inc. · Aladdin risk analytics and neural-network development
Aladdin Risk analytics, scenario analysis and stress testing exist. The precise claim that deployed ML improves those core models at scale is not established; neural-network pricing/risk work is described as development. Avoid substituting the separately cataloged Aladdin Copilot.
Lead sourceblackrock.com · opens in a new tab
Lead sourceengineering.blackrock.com · opens in a new tab
Ant Group Co., Ltd. · Zhima/Sesame online-data credit scoring
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcealizila.com · opens in a new tab
Lead sourcealibabagroup.com · opens in a new tab
Allstate Corporation · AI Pricing Optimization — deployment unverified
Primary project account contradicts the cited workflow and does not establish production pricing, competitive-pricing inputs, retention optimization or deployment.
Lead sourcebreakthroughtech.org · opens in a new tab
Ant Group Co., Ltd. · Ant Group AI Fraud Prevention — entity and scope unverified
Wrong entity and workflow in the supporting citation; do not transfer Ant International's claims to Ant Group or Alipay.
Lead sourcebusinesswire.com · opens in a new tab
BlackRock, Inc. · AI-Driven ESG Scoring — distinct initiative unverified
General systematic NLP evidence supports the technology but not the precise proprietary ESG scoring claim. Potential overlap with the existing Systematic Active Equity AI Signals record requires adjudication.
Lead sourceblackrock.com · opens in a new tab
Lead sourcecbinsights.com · opens in a new tab
Citadel LLC · NLP for Market Sentiment Analysis — scope unverified
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcecitadelsecurities.com · opens in a new tab
Citigroup Inc. · AI Risk Model Validation — deployment unverified
A general discussion of industry risk and compliance uses is not evidence for this exact Citi deployment.
Lead sourcecitigroup.com · opens in a new tab
Edward Jones Investments · AI financial-planning investments — deployment unverified
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourceedwardjones.com · opens in a new tab
Goldman Sachs Group, Inc. · AI-Powered Deal Preparation — exact workflow unverified
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcepymnts.com · opens in a new tab
JPMorgan Chase & Co. · DocLLM research for layout-aware document understanding
Research and AI architecture are supported; the Production label and 2024-01-01 announcement date are not. Hold as research evidence.
Lead sourcearxiv.org · opens in a new tab
Nomura Holdings, Inc. · AI Risk Analytics Platform — entity and workflow unverified
NRI and Nomura Holdings are different entities. The cited 2024 announcement cannot substantiate this institution's claimed production workflow.
Lead sourcenri.com · opens in a new tab
The Charles Schwab Corporation · AI Portfolio Insights for Advisors — unverified legacy claim
The primary launch describes a retail-client informational product that disclaims investment recommendations. It does not establish the original advisor-facing portfolio recommendation system; source correction cannot establish shared initiative identity.
Lead sourcepressroom.aboutschwab.com · opens in a new tab
The Progressive Corporation · AI Fraud Detection in Claims — exact workflow unverified
Broad issuer AI/fraud statement supports only a general capability; exact claims workflow and operational outcomes remain unverified.
Lead sourceprogressive.com · opens in a new tab
Two Sigma Investments, LP · NLP for Alternative Data Processing — duplicate consolidation
Catalog-level consolidation: the primary account describes NLP as one technique in investment-signal modeling, without a separate named deployment. Attach supported evidence to ML-Driven Investment Signals and retain the unsupported feed-specific claims in this audit; issuer-confirmed identity is not asserted.
Lead sourcetwosigma.com · opens in a new tab
Wells Fargo & Company · AI Data Quality Management — exact workflow unverified
Responsible-AI governance is not evidence for the specific automated data-cleansing workflow.
Lead sourcestories.wf.com · opens in a new tab
Goldman Sachs Group, Inc. · AI Equity Research Summarization
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcegoldmansachs.com · opens in a new tab
Morgan Stanley · AI Document Summarization for IB
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceopenai.com · opens in a new tab
HSBC Holdings plc · Dynamic Risk Assessment financial-crime monitoring
Legacy description incorrectly broadens AML to credit, market and operational risk. This is the same Google AML initiative as cmmpo6igf004jvwfgd2p82j6l; consolidate evidence, preserving both originals.
Lead sourcehsbc.com · opens in a new tab
Barclays PLC · AI-Powered Credit Decisioning
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcestatic.seekingalpha.com · opens in a new tab
The Charles Schwab Corporation · Schwab Intelligent Portfolios
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceschwab.com · opens in a new tab
Fidelity Investments · AI Stock Selection Models
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceinstitutional.fidelity.com · opens in a new tab
The Vanguard Group, Inc. · Vanguard Digital Advisor
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceinvestor.vanguard.com · opens in a new tab
Lead sourceownyourfuture.vanguard.com · opens in a new tab
Citadel LLC · AI Risk Management Systems
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcecitadel.com · opens in a new tab
Two Sigma Investments, LP · AI Portfolio Optimization
Research supports an AI method, not the asserted production position-sizing and transaction-cost optimization system.
Lead sourcetwosigma.com · opens in a new tab
JPMorgan Chase & Co. · AI Threat Modeling Co-Pilot for software security design
Source concerns design-time threat modeling, not live network/log anomaly detection. It overlaps existing Auspex threat-model generation research_601d913632d634840ef6324d; resolve same-initiative identity before merging.
Lead sourcejpmorganchase.com · opens in a new tab
Citigroup Inc. · Citigroup AI Cybersecurity Operations
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcecitigroup.com · opens in a new tab
UBS Group AG · UBS AI Regulatory Change Management
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceubs.com · opens in a new tab
JPMorgan Chase & Co. · AI-Driven AML/KYC Screening
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcejpmorganchase.com · opens in a new tab
Goldman Sachs Group, Inc. · Historical Marcus credit decision model — AI basis unverified
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcegoldmansachs.com · opens in a new tab
Lead sourcegoldmansachs.com · opens in a new tab
Bank of America Corporation · AI Credit Decisioning
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcenewsroom.bankofamerica.com · opens in a new tab
UBS Group AG · UBS AI portfolio rebalancing — exact claim unverified
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceubs.com · opens in a new tab
Citigroup Inc. · Exploratory AI payment-routing work
The original source is a June 2024 cross-border-payments outlook video. Citi's payment-AI whitepaper explicitly describes routing tools as exploration, not a production implementation.
Lead sourcecitigroup.com · opens in a new tab
Lead sourcecitigroup.com · opens in a new tab
Wells Fargo & Company · Explainable lending-risk model research — production mortgage claim unverified
A historical explainable-AI research presentation is identified; the original link now redirects and primary full text was not recovered. Hold as a research lead, not a production mortgage deployment.
Lead sourcenvidianews.nvidia.com · opens in a new tab
Lead sourceuk.marketscreener.com · opens in a new tab
Deutsche Bank AG · AI-Driven FX Trading
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcedb.com · opens in a new tab
Fidelity Investments · Fidelity AMP automated managed accounts — AI basis unverified
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourceir.53.com · opens in a new tab
The Vanguard Group, Inc. · AI Index Fund Optimization
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcecorporate.vanguard.com · opens in a new tab
BNP Paribas S.A. · AI Trade Surveillance System
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcegroup.bnpparibas · opens in a new tab
Lead sourcegroup.bnpparibas · opens in a new tab
Crédit Agricole regional banks (Brittany cooperation) · Scoring.AI customer targeting at Crédit Agricole regional banks
A vendor case supports a narrower regional-bank workflow. It does not establish deployment at the listed Crédit Agricole S.A. entity or across its entire retail network; hold pending institution resolution.
Lead sourcetalan.com · opens in a new tab
Morgan Stanley · Morgan Stanley AI Compliance Monitoring
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcemorganstanley.com · opens in a new tab
Wells Fargo & Company · AI/ML in financial-crime risk management — precise system unverified
Retain as a lead pending workflow-level primary deployment evidence; do not certify millions of transactions, real-time detection or measured false-positive reductions.
Lead sourcebankingjournal.aba.com · opens in a new tab
Lead sourcewellsfargojobs.com · opens in a new tab
Nomura Holdings, Inc. · Nomura FX execution algorithms — AI basis unconfirmed
Unsupported AI basis: technology is explicitly unknown, absent or non-AI; retain as a lead.
Lead sourcenomuraconnects.com · opens in a new tab
BNP Paribas S.A. · Operational GenAI pilot — specific workflows unverified
Hold the precise legacy claim; broad GenAI adoption cannot validate three asserted workflows or a January 2024 start.
Lead sourcecib.bnpparibas · opens in a new tab
Lead sourcegroup.bnpparibas · opens in a new tab
Aetram Trades · Interest in MetaTrader AI strategy coding
Avoid attributing a vendor capability to a broker without adoption evidence.
Lead sourcemetaquotes.net · opens in a new tab
Amundi S.A. · Generative AI scenario and allocation analysis update
Strong new outcome disclosure, held for exact initiative identity rather than splitting a potentially existing workflow.
Lead sourceglobenewswire.com · opens in a new tab
Lead sourceresearch-center.amundi.com · opens in a new tab
Jiufang Intelligent Holdings Limited · Jiufang Lingxi research assistant on Huawei Xiaoyi
Specific in-window deployment lead held for primary corroboration.
Lead sourceclient.sina.com.cn · opens in a new tab
Noor Capital · Planned test of MetaTrader AI Assistant
Retain as an announced test plan only.
Lead sourcemetaquotes.net · opens in a new tab
RiskX · Planned personalized structured-investment AI proof of concept
Aspirational MOU retained separately from deployed evidence.
Lead sourceglobenewswire.com · opens in a new tab
AmericanFortress · Reusable crypto identity credentials linked to wallet names
Future availability and no explicit AI mechanism for the workflow.
Lead sourcesumsub.com · opens in a new tab
Cashfree Payments · AI-agent assisted hosted checkout with customer payment approval
Reviewed: Independent source audit: WebMCP exposes deterministic existing checkout actions to external browser agents. No observed agent/customer execution or deployed AI decision workflow is established; retain infrastructure availability as a lead.
Lead sourcecashfree.com · opens in a new tab
Lead sourceexpresscomputer.in · opens in a new tab
Huntington Bancshares Incorporated · Multi-bank cash visibility and forecasting platform
Intelligent branding and a Kyriba relationship do not establish AI use.
Lead sourceir.huntington.com · opens in a new tab
Jelou AI · Conversational onboarding with identity verification
Named partnership and explicit AI, but joint production or pilot evidence is missing.
Lead sourcesumsub.com · opens in a new tab
Klarna · Global in-app caller verification
Deployment is verified but the release does not attribute the feature to AI; general AI network boilerplate is insufficient.
Lead sourceinvestors.klarna.com · opens in a new tab
Natural · Payment-rail selection interfaces for AI agents
Infrastructure is live, but a production AI decision workflow is not established.
Lead sourcenatural.com · opens in a new tab
Payment Asia · Merchant health intelligence and AI marketing launch
Explicit AI launch, but no confirmed active pilot cohort or production usage.
Lead sourcepaymentasia.com · opens in a new tab
Singapore Police Force · Joint cryptocurrency scam intervention with payment token providers
Primary deployment evidence exists, but AI is not explicit; avoided attributing all prevented losses to AI.
Lead sourcepolice.gov.sg · opens in a new tab
Chimo AI Inc. · AI-assisted matching to unclaimed funds and settlements
Explicit AI claim but weak deployment provenance and no independently verified outcome.
Lead sourcechimo.ai · opens in a new tab
Lead sourcechimo.ai · opens in a new tab
Sovos Compliance, LLC · Unclaimed-property due diligence and reporting automation
Automation is established; broad Sovos AI branding does not prove AI in this product.
Lead sourcesovos.com · opens in a new tab
Lead sourceup.sovos.com · opens in a new tab
Stripe, Inc. · Quarterly inactive connected-account review and escheatment outreach
Only indexed or blocked sources; substantive evidence access remains unverified.
Lead sourcesupport.stripe.com · opens in a new tab
Lead sourcesupport.stripe.com · opens in a new tab
Texas Comptroller of Public Accounts · Announced AI modernization of unclaimed-property claims
Genuine in-window announcement; active AI claims processing, vendor and technical scope not established.
Lead sourcecomptroller.texas.gov · opens in a new tab
Utah Office of State Treasurer · AI call notes in unclaimed-property claims workflows
Explicit AI and relevant workflow, but date and deployment status insufficiently established.
Lead sourcele.utah.gov · opens in a new tab

Read the research briefing