Worrying open-source security issue 'BadHost' could affect millions of AI agents, experts warn - TechRadar
9/10Experts warn that the open-source security vulnerability 'BadHost' discovered in May 2026 could affect millions of AI agents worldwide, posing serious security risks by potentially allowing malicious actors to compromise or control these systems. This issue stresses the critical need for securing open-source components within AI agent deployments.
