Cost savings · Cited evidence brief
AI classifies customer-service case logs
$4.9m
reported classification savings over twelve months
Twelve months ended June 30, 2025
A measurement the organization reports in its own filing, report or release. Not independently audited.
Reported facts · from the source
- Value
- $4.9m
- What was measured
- reported classification savings over twelve months
- Period
- Twelve months ended June 30, 2025
- Baseline
- Not stated in the reviewed source summary
- Source published
- Sep 10, 2025
- Scope
- Internal case-log classification; historical twelve-month savings, with a separate August 2024 internal accuracy test.
- Evidence kind
- Company filing
$4.9m
reported classification savings over twelve months
Twelve months ended June 30, 2025
Scope, baseline & source
- Scope / denominator
- Internal case-log classification; historical twelve-month savings, with a separate August 2024 internal accuracy test.
- Baseline
- Not disclosed in this account
- Claim basis
- Company filing
Savings and internal accuracy refer to different periods. Class balance, test design and attribution methodology are undisclosed. Primary filing text was reviewed through indexed content after full extraction hit a size limit. This is distinct from the support-assistant savings and should not be added without checking accounting overlap.
Read claim source ↗Measures retain their original currency, scope and period. Capacity, adoption and forecasts are not realized cash savings.
Documented vendor relationships
No vendor relationship has been linked to this specific workflow in the curated overlay. This does not mean it was built without vendors.
See the company’s documented relationshipsWhat the source reports
Klarna reports classifying over one million monthly chat conversations with AI and approximately USD4.9 million in savings over the twelve months ended June 30, 2025. An internal August 2024 accuracy test reported 96%.
How to read this number
Savings and internal accuracy refer to different periods. Class balance, test design and attribution methodology are undisclosed. Primary filing text was reviewed through indexed content after full extraction hit a size limit. This is distinct from the support-assistant savings and should not be added without checking accounting overlap.
Human controls and constraints
Stated in the source summary or the editorial caveat.
- Accuracy and savings have different evaluation periods
- Per-class error rates not disclosed
- Potential accounting overlap prevents summing savings
Brief reviewed Oct 26, 2026 against the primary source above. The review date is separate from the source publication date and measurement period. Not independently audited.