Can you trust your AI? 7 questions to ask your solution vendor

How do you trust your AI solution to manage your AML risk?

AI solutions are now used in one third of financial institutions for Anti-Money Laundering purposes. Another third of institutions are currently experimenting with pilot projects¹.

And yet ensuring we can trust AI-based solutions is a major challenge for AML teams who are looking to deliver effective and efficient new approaches.

Risk-based approaches are used across financial institutions, but the introduction of AI solutions presents new challenges for managing risk. The question of trust in AI is one of the most important topics when financial institutions are considering whether or not to invest or indeed widely roll out such solutions.

Our own experiences building AML investigation solutions for Tier 1 global banks have helped us to develop 7 questions that we think you should be asking your own AI solution vendor to better understand how they manage risk and ensure solutions can be trusted.

The questions are listed below but our full list also includes context and examples of the answers you should be looking out for. This is by no means exhaustive but they’re definitely some of the top subjects that any effective AML or risk team should be thinking about.

If you’re interested in a deeper analysis of how to use a risk-based approach to manage AI-based solutions then read our paper here.

7 questions to ask your AI solution vendor:

1. What are the major risks with your AI solution?
2. What is the performance of your AI solution on the high-risk and high-impact classes?
3. How do you ensure the quality of your training data for your AI solution?
4. How do you detect if your AI solution decisions are biased?
5. How do you know if live cases are outside your AI solution safe limits?
6. How can humans review your AI solutions outputs effectively and efficiently?
7. How do you resolve contested and wrong decisions from your AI solutions?

Further reading

For a deeper understanding of the subject of trust in AI, you can find our paper ‘Trusting ML in Anti Money Laundering: A risk-based approach’ here.

(1) Institute of International Finance, (2018). Machine Learning in Anti-Money Laundering (www.iif.com/Publications/ID/1421/Machine-Learning-in-Anti-Money-Laundering)