Key Takeaways:

• Using AI and predictive capability is not new but generative AI brings AI to the mainstream due to its human-like communication and ease of accessibility to companies of all sizes.
• For boards to properly oversee AI risk, they need to require management to establish a model risk management process along with a model risk governance framework around AI implementation that includes standardization of algorithms, governance and ethical considerations as well as validation of data inputs and outputs.
• Moving to AI changes the human role and underscores the need for a strong risk culture and exercising of healthy skepticism to contend with potential bias, complexity and lack of integrity within both data sets and AI algorithms.
• Underlying data sets must be of high quality and representative of population they are intending to serve – free from bias, properly labeled and annotated.
• Boards, management and auditors will need to continue to stay apprised of what is sure to be a continually evolving standard setting and rule-making environment as technologies continue to evolve.