From Monitoring to Automation: How AI is Redefining Financial Crime Compliance

AI is no longer just an experimental tool in compliance - it’s rapidly becoming the backbone.

Earlier this week I met with some of the leading practitioners from the financial crime compliance (FCC) departments in Hong Kong banking sector.

And here is what I learnt.

AI Moves from Pilot to Production

Banks like HSBC Hong Kong and Standard Chartered are already using AI to automate major chunks of transaction monitoring and internal investigations.

The result?

Some estimate that 50-60% of manual FCC tasks have already been automated, freeing human teams to focus on judgment-heavy cases.

Compliance people are worried (as they should be)

On the job security front, my most intuitive guess is that AI won’t replace compliance officers. At least not in the near future. But compliance officers using AI will replace those who don’t.

Regulators Are (Carefully) Opening the Door

In the meantime, the Hong Kong Monetary Authority (HKMA) and the Monetary Authority of Singapore (MAS) are encouraging - not mandating - AI use.

• Banks submitted their AI application reports to the HKMA last month.

• The regulator is now studying the results before issuing formal guidance.

• The document should be issued later this year.

In the meantime, the approach is one of caution. Use AI but do it responsibly.

Responsibly?

The explainability problem.

Here’s the catch: regulators, investigators, and even internal compliance teams demand clarity. If AI can’t explain why it flagged a transaction or cleared a client, it’s a non-starter.

Explainability isn’t a nice-to-have anymore - it’s a regulatory expectation.

What's Next?

Compliance officers and IT teams must get more hands-on with AI - not just at the model level, but understanding the risks and governance around it.

Expect more pressure from regulators for clear explanations, strong data governance, and internal accountability when using AI in FCC.

In a field where speed and precision matter, those who adapt fast (and responsibly) will lead.

Thanks for reading.
Have suggestions, questions, or examples from your own experience?
Reach out on LinkedIn or email - I am always interested in fresh insights.