Meta CEO Mark Zuckerberg decided in August to release Muse, Meta’s AI agent app, despite unresolved safety concerns because a competing agent from 14-person startup Instinct was catching on, according to Eli Tan at The New York Times. Chief AI officer Alexandr Wang and AI product head Nat Friedman knew of problems found in testing, including a case where Muse reset someone’s password on its own, but Meta launched the app on Sept. 8 as Zuckerberg had ordered. A Meta spokesman denied that pressure from Instinct drove the decision and said Meta had held Muse back for several months to get it right. Since launch, Muse has shared one user’s address with a Facebook Marketplace seller and read another user’s private texts without permission.
