Skip to main content

Treat AI models as would-be hackers, says quant

Risk Live: Models capable of “strategic deception” require different risk management, says former Risk.net quant of the year

Finger presses warning alert sign for security system or software

Banks should keep watch over frontier artificial intelligence models as much as they look out for rogue traders, fraudsters and hackers, a leading quant says.

Alexander Sokol, founder of risk management technology company CompatibL, said agentic AI presents a new type of risk in which the adversary – though a computer model – hides its behaviour, lies and disrupts efforts to police it. Sokol was speaking at Risk Live Europe in London on June 30.

The new “adversarial model risk” is starkly

Sorry, our subscription options are not loading right now

Please try again later. Get in touch with our customer services team if this issue persists.

New to Energy Risk? Register here

Most read articles loading...

You need to sign in to use this feature. If you don’t have a FX Markets account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: