# Commonwealth Bank's AI fraud detection: 20M+ payments scanned daily, fraud losses down over 20%
> In its February 2026 'Our Approach to Adopting AI' report, Commonwealth Bank of Australia said AI-driven scam and fraud detection scans more than 20 million payments a day and sends 40,355 proactive warning alerts daily, and that this played a crucial role in reducing customer fraud losses by over 20% in 1H26 versus 1H25. Its Compass AI assistant has answered more than 500,000 questions since July 2024. Every figure is CBA's own, self-reported in its AI-adoption report.
- Verification status: pending
- Case type: deployment
- Provider: Commonwealth Bank of Australia, in-house AI (https://www.commbank.com.au/)
- Client: Commonwealth Bank of Australia (CBA), Banking (Australia) (named)
- Sector: banking / Australia / finance
- Canonical URL: https://theinternetninja.com/stories/commbank-cba-ai-fraud-detection/
- Source: The Internet Ninja (theinternetninja.com), independent verified-proof platform
## Outcomes
| Metric | Before | After |
| --- | --- | --- |
| Payments scanned by AI fraud detection | | more than 20 million daily on average |
| Customer fraud losses, 1H26 vs 1H25 | 1H25 baseline | reduced by over 20% |
| Proactive scam/fraud warning alerts to customers | | 40,355 per day on average, via the CommBank app |
| Compass AI questions answered since July 2024 | | more than 500,000; inquiries delivered three times faster than traditional methods |
## Verification method
Sourced by The Internet Ninja against the public record: every figure below is quoted from the cited source and, on 2026-07-08, checked against the live page over the open network, with a web.archive.org capture taken the same day. This case is pending, it has not yet cleared TIN's green validation bar. The single source is CBA's own newsroom announcement of its February 2026 'Our Approach to Adopting AI' report: the figures are first-party and self-reported by the bank about its own systems, not independently audited, and CBA is both the provider and the client here. The over-20% fraud-loss reduction is stated by CBA as one crucial contributor, not an isolated, attributed measurement. A green badge requires independent corroboration of the fraud-loss reduction and the Compass figures, or CBA confirming the measurement basis on the record.
## FAQ
**What did Commonwealth Bank report about its AI fraud detection?**
Per CBA's February 2026 AI-adoption report, AI-driven scam and fraud detection scans more than 20 million payments a day and sends 40,355 proactive warning alerts daily, and CBA says this played a crucial role in reducing customer fraud losses by over 20% in the first half of its 2026 financial year versus the same period a year earlier.
**Why is this case file pending, not verified?**
Every figure comes from CBA's own report about its own systems, self-reported and not independently audited, and CBA is both provider and client. A green badge requires independent corroboration or CBA confirming the measurement basis on the record.
## Full case file
## The problem
Payment scams and fraud move at machine speed and volume: a retail bank clears tens of millions of transactions a day, and a human team cannot inspect each one for the patterns that signal a scam in progress. Per its February 2026 report, Commonwealth Bank of Australia applies AI to that scale problem across its payment flows and its frontline banker tooling.
## What was built
CBA describes AI methods "applied to process more than 20 million payments daily on average and send 40,355 proactive warning alerts per day on average to customers via the CommBank app" ([source](https://www.commbank.com.au/articles/newsroom/2026/02/cba-approach-to-adopting-ai-report-announcement.html)). Alongside fraud detection, its generative-AI assistant for business bankers, Compass AI, "delivers inquiries from its business bank knowledge base three times faster than traditional methods and has been used to answer more than 500,000 questions since July 2024" ([source](https://www.commbank.com.au/articles/newsroom/2026/02/cba-approach-to-adopting-ai-report-announcement.html)).
## The outcome
**Fraud losses down over 20%.** Per CBA, the AI-driven detection "has played a crucial role in helping to reduce customer fraud losses by over 20% in the first half of 2026 financial year compared to the first half of 2025" ([source](https://www.commbank.com.au/articles/newsroom/2026/02/cba-approach-to-adopting-ai-report-announcement.html)). CBA frames the AI as a crucial contributor to that reduction, not the sole cause, and this file reports it that way.
**Compass AI, at the frontline.** The 500,000-plus questions answered since July 2024 and the three-times-faster figure are CBA's, describing how the assistant lets business bankers answer complex customer questions from the bank's knowledge base ([source](https://www.commbank.com.au/articles/newsroom/2026/02/cba-approach-to-adopting-ai-report-announcement.html)).
## How this was verified
This case is pending, and the caveat is structural: every figure comes from CBA's own newsroom announcement of its own AI-adoption report. The bank is both the provider and the client, the figures are self-reported and not independently audited, and the fraud-loss reduction is stated as something the AI helped drive rather than an isolated, attributed measurement. What keeps it on the record is that the source is a named, dated, first-party report with specific figures, which makes the claims checkable and falsifiable. Green requires independent corroboration of the fraud-loss and Compass figures, or CBA confirming the measurement basis on the record.
---
## Sources
Cited in this case file. Tier 1 = first-party primary (the company's own report). Each figure was checked against the live source on 2026-07-08.
1. Commonwealth Bank newsroom, "CBA's approach to adopting AI" (announcing the February 2026 'Our Approach to Adopting AI' report; Tier 1, first-party, self-reported). ยท [archived](https://web.archive.org/web/20260708222642/https://www.commbank.com.au/articles/newsroom/2026/02/cba-approach-to-adopting-ai-report-announcement.html)