AI start-up sees thousands of vulnerabilities in popular tools – Washington Post

Haize Labs hopes to become a “Moody’s for AI” that establishes public-safety ratings for popular artificial intelligence models.
Source: Google News Updates