“Show me, don’t tell me”: How AI is transforming compliance

Regulators, boards and senior leaders increasingly expect firms to demonstrate that compliance controls work, not just that they exist, and TCC argues predictive AI can provide that continuous evidence.

What happened?

TCC’s analysis describes a shift in regulatory expectations: firms are no longer able to rely on asserting that compliance policies exist. The FCA, boards and senior leaders now expect clear, data-backed evidence that those policies are working in practice.

The piece sets out how AI tools can analyse large volumes of customer interactions and operational data to identify patterns, highlight risks and generate real-time evidence of conduct, in place of retrospective reviews or limited manual sampling.

TCC and Recordsure hosted a related live webinar on 6 November 2025, exploring how predictive AI can improve monitoring, streamline assurance processes and build regulatory trust without increasing headcount.

Why does it matter?

Traditional oversight methods struggle to keep pace with rising volumes of customer interactions, data and regulatory expectations. Firms that rely solely on retrospective reviews or small manual samples risk gaps between what they report and what is actually happening in customer-facing conduct.

Predictive AI is presented as a way to automate repetitive monitoring tasks, freeing compliance teams to focus on strategic oversight and to surface insights that would otherwise take weeks of manual analysis.

Who is affected?

Compliance, risk and assurance functions across financial services firms, particularly those under pressure to evidence oversight without expanding headcount.

Key risks

  • Compliance assurance based on assertion rather than demonstrable, data-backed evidence.
  • Retrospective reviews and limited manual sampling that cannot keep pace with the volume of customer interactions.
  • Slow identification of emerging conduct risks when analysis depends on manual processes.

Actions to take

  1. Assess whether current monitoring can demonstrate, not just assert, that controls are working.
  2. Identify where AI-based analysis could replace retrospective or sample-based reviews.
  3. Register for TCC and Recordsure’s live webinar on AI-led compliance assurance.

Wider implications

As “show me, don’t tell me” scrutiny becomes the norm, firms able to produce continuous, objective evidence of conduct are better placed to respond to regulator, board and stakeholder questions than those relying on periodic reporting alone.

Recommendations

Firms should review where compliance assurance still depends on manual sampling or retrospective checks, and consider where predictive AI, used alongside regulatory expertise, could provide continuous evidence of good conduct.

Supporting sources

  1. “Show me, don’t tell me”: How AI Is transforming compliance

Frequently asked questions

What does 'show me, don't tell me' compliance mean?

It means firms must provide data-backed evidence that compliance controls are working, rather than simply stating that policies are in place.

How can AI help with compliance assurance?

AI can analyse large volumes of customer interactions and operational data to identify patterns, highlight risks and provide real-time evidence of conduct.

When is TCC and Recordsure's related webinar?

The live webinar took place on 6 November 2025 at 12:30pm.

Ready to strengthen your compliance?

Speak to our experts about your regulatory challenges.