Financial Crime in 2026: Better Tools, Harder Judgement Calls

Financial crime controls are entering a period in which institutions have stronger information-sharing powers and better analytical tools, but success still depends on judgement, curiosity and accountable decisions.

A review of the Institute’s 2026 financial crime conference highlighted UK reforms that allow firms to share warnings and requests where economic crime concerns arise.

One pilot reportedly enabled a warning about a network of accounts to trigger investigations across ten institutions, leading to £1.4 million in frozen funds and the closure or rejection of numerous relationships.

Technology cannot replace plausibility checks

The same discussion stressed that compliance teams must ask whether a customer’s explanation makes commercial sense. Income projections, commissions and source-of-wealth claims should be compared with real market conditions rather than copied into a file without challenge.

AI can help connect data and prioritise alerts, but it remains a tool rather than a decision-maker. High-risk onboarding and enhanced due diligence still require local context, clear reasoning and human accountability.

The direction of travel is encouraging: firms can share more and analyse faster. The harder task is building a culture that uses those powers consistently and records why action was or was not taken.

Information sharing needs disciplined governance

New legal permissions do not mean every suspicion should be distributed widely. A warning or request should be necessary, proportionate and supported by an auditable rationale. Firms need controls over who can send information, which data fields are shared and how the recipient may use them.

Shared intelligence should feed into an investigation rather than dictate an automatic exit. Each institution remains responsible for assessing its own customer and avoiding unfair outcomes. Governance should therefore combine speed with data quality, challenge and the ability to correct information that later proves inaccurate.

Questions for compliance leadership

• Can investigators obtain relevant information from peer institutions quickly?
• Are customer explanations tested against external commercial reality?
• Do AI tools provide reasons that reviewers can challenge?
• Is every high-risk decision supported by a clear written rationale?