Anthropic is testing AI’s capacity for sabotage- whether AI could conceal its true capabilities or bypass safety mechanisms October 22, 2024 // by Finnovate This content is for members only. Sign up for access to the latest trends and innovations in fintech. View subscription plans. Login