Artificial Intelligence applied to Risk Management

The paper Artificial Intelligence applied to Risk Management aims to guide risk managers on applying AI. It goes from basic learning to identifying new risks and developing their own strategy on how to take the next steps regarding the implementation of AI. The basics of AI are explained with an emphasis on its requirements and its limits.

As with every new technology, there are emerging risks and challenges related to AI. Questions on
ethics, equality, biases, trustworthiness and reliability of decisions made or suggested by AI, are just
some of the topics that need attention. Transparency of algorithms and increasing cyber risks are also
issues for the risk manager to address.

Some strategic questions that organisations need to address include:

• How and why is our company using and applying AI?
• What new liability or cyber challenges arise?
• What are the challenges for our workforce on decision, on ethics and biases?

In the AI race, not only companies and scientists are competing with each other, but also countries and
governments, because the related impact goes beyond companies and individuals; today’s research
and decisions will affect societies. Europe has the opportunity to define its own way on how to apply AI
and handle the related risks and fears in what the European Parliament described once as a “humancentric”
approach to AI.

With this document, FERMA aims to kick off a broader discussion on AI and risk management.