Artificial intelligence has become the opening topic of almost every anti-money laundering (AML) discussion this year, but RegTech firm Napier AI argues that the race to adopt it is leaving many compliance teams exposed. According to the company, AI is not a shortcut to strong AML compliance. It can only perform as well as the foundations it is built on.

Napier AI notes that vendors are promoting AI across the board, roadmaps are full of it, and compliance functions feel pressure to act quickly. The problem, it says, is that too many programmes begin with the technology rather than the issue it is supposed to address. Firms often add a new model to an existing process to reduce alert volumes or speed up investigations without first checking whether the underlying controls are sound. If they are not, AI will not repair them. Instead, it makes those weaknesses harder to spot.

One misconception Napier AI wants to dispel is that AI exists to eliminate false positives. A system generating no alerts at all is not efficient, the firm argues; it is more likely accepting risk it should not. The more useful question is whether the right alerts are being produced, and whether they can be prioritised, reviewed and evidenced in line with the organisation’s risk appetite. Regulators including the Financial Conduct Authority (FCA) in the UK and the Monetary Authority of Singapore (MAS) consistently expect firms to understand their risks first and then apply proportionate controls.

That is why Napier AI sees risk appetite as the true foundation.

Source link