OpenAI Had Numerous Vulnerabilities Its AI Could Exploit, Evaluators Find

T1Covered by 1 source · first reported by The Information (teasers) at 22 Oct, 15:00 UTC

OpenAI’s systems had multiple vulnerabilities that would allow AI agents to circumvent monitoring and take arbitrary unauthorized actions, evaluators from AI safety nonprofits METR and Redwood Research found during a three-week pilot with OpenAI from late July to early August. The nonprofits ...

aisecurity

Sources

• T1OpenAI Had Numerous Vulnerabilities Its AI Could Exploit, Evaluators FindThe Information (teasers)22 Oct, 15:00
• T1OpenAI Had Numerous Vulnerabilities Its AI Could Exploit, Evaluators FindThe Information (teasers)22 Oct, 15:00