New Reports Uncover Jailbreaks, Unsafe Code, and Data Theft Risks in Leading AI Systems
Multiple AI jailbreaks and tool poisoning flaws expose GenAI systems like GPT-4.1 and MCP to critical security risks.
More info