Peter Slattery, Neil Thompson, Alexander Saeri and Emily Grundy, have released the first-ever AI Risk Repository: a comprehensive and accessible living database of 700+ risks posed by AI that will be continuously updated to ensure relevancy and timeliness.
Working in collaboration with Michael Noetel, Risto Uuk, Soroush Pour, James Dao, and Stephen Casper, from the University of Queensland, Future of Life Institute, KU Leuven, and Harmony Intelligence, and MIT CSAIL.
For a comprehensive overview of their research and insights, you can access the website here.
The release of the paper was accompanied by media coverage, including:
Techcrunch: MIT researchers release a repository of AI risks
ZDNET: AI risks are everywhere - and now MIT is adding them all to one database
VentureBeat: MIT releases comprehensive database of AI risks
MIT Technology Review: A new public database lists all the ways AI could go wrong
Wired: Researchers Have Ranked AI Models Based on Risk—and Found a Wild Range
Observer: MIT Launches the First Ever Comprehensive Database of A.I.