Skip to content Skip to footer

Loading Results

Model Risk Management of AI and Machine Learning Systems

The purpose of this document is to present a model risk management approach for applied artificial intelligence systems. It reflects the nascent AI regulatory landscape and its expected near term development. AI systems are found to be consistent with broader definitions of 'models' applied in the financial industry, e.g. in regulatory guidance on model risk management, and therefore the latter are perceived as applicable, provided that extensions are made that address AI-specific risks. In this paper we review in detail risks specific to AI, recommended approaches to deal with these risks, and the necessary adjustments to an existing model risk management framework.

Contact us

Matthew Dodgson

Director, UK LIBOR Models Team, PwC United Kingdom

Tel: +44 (0)7801 766052

Fabrice Ciais

Associate Director, AI, PwC United Kingdom

Tel: +44 (0)7843 334241

Follow us