Visa has launched six AI tools to automate charge dispute processing, targeting a problem that has exploded: 106 million disputes processed in 2025, up 35% since 2019. For UK financial services firms, insurers and their compliance teams, this matters because disputes touching consumer protection are explicitly in scope of FCA Consumer Duty PS22/9. Visa's approach—using generative AI to detect and respond to disputes before escalation—is a blunt signal that manual dispute handling is no longer fit for purpose. The tools flag suspicious patterns and generate responses automatically. The question is not whether Visa's system works; the question is whether your firm is still doing this by hand.

What Visa's move reveals is a wider pattern: payment networks and platform operators are moving faster on regulated workflow automation than the firms they serve. They have the data scale and technical resources to train AI on millions of real disputes. Most UK law firms, accountancy practices and mid-market financial services firms do not. This creates an asymmetry. Visa standardises dispute handling at the network level; individual regulated firms remain fragmented, under-resourced and exposed to error. Generative AI has made this gap visible. Tools like Microsoft Copilot and Harvey promise to close it, but they work best on high-volume, well-structured work. Disputes are structured. The problem is that most firms have not yet built the data governance foundation those tools need—and regulators under PRA SS1/23 and the draft EU AI Act are watching.

Here is what Trovix sees most firms get wrong about AI in dispute workflows: they buy a tool and expect it to work immediately. Visa can do that because Visa owns the infrastructure. Your firm cannot. Generative AI needs governance before it needs capability. It needs Trovix Audit to track decisions, flag drift, and prove to the FCA that your AI is auditable. It needs data hygiene—standardised dispute schemas, clean historical data, clear ownership of inputs and outputs. Without that, you risk deploying AI that looks smart but fails under regulatory scrutiny. Generative AI can absolutely help you detect patterns and draft responses faster than humans. But it cannot replace human judgment on consumer complaints that touch redress and fair conduct. The gap between what Visa's tools do and what your firm needs to do is not about the AI. It is about the governance wrapper around the AI.

If you handle charge disputes, chargeback disputes or payment disputes, this is the moment to act. Run an honest audit of your current process: who touches each dispute, what decisions are made, how are exceptions escalated, who owns the consumer experience? Then map where AI could reduce human time on pattern-detection and routine responses—without removing human sign-off on decisions that affect customers. Trovix Brief can help you automate intake and triage so disputes reach the right person faster. But do not rush to generative AI without first building the compliance and audit foundations. Visa's move is a signal that the industry is moving. Waiting another two years to act is a choice—but not a smart one.

Source: CNBC