You’re reading the web edition of STAT’s AI Prognosis newsletter, our subscriber-exclusive guide to artificial intelligence in health care and medicine. Sign up to get it delivered in your inbox every Wednesday.
Omada’s threat model for AI
I’m clumsy, which made for the occasional broken flask when I worked in a chemistry lab. I eventually realized “I hope I don’t knock that over” really meant “I should move that so I can’t knock it over.” The same thing applies to AI: If you don’t want your AI agent to do something like bite you, don’t give it teeth.
Virtual care provider Omada recently released PROMISE TO MAP, a “threat model” that helps healthcare organizations assess the security, privacy, and compliance risks introduced by using artificial intelligence. Among topics the framework covers is how basic data governance and cybersecurity principles apply to AI — something that’s easy to forget in the scramble to contain a new technology that feels alien.
This article is exclusive to STAT+ subscribers
Unlock this article — plus in-depth analysis, newsletters, premium events, and news alerts.
Already have an account? Log in
