Delve, a startup handling AI security certifications, itself became the weak link in Context AI's data breach

Context AI, an AI agent training startup, disclosed a security incident last week. TechCrunch confirmed that Delve — a compliance company that certifies whether companies meet security standards — was the firm that performed Context AI's security certifications before the breach occurred.

This means a company trusted to verify that other AI startups are secure failed to catch vulnerabilities in at least one of its own customers, raising questions about how thorough these third-party security audits actually are in the fast-moving AI industry.

For business leaders evaluating AI startups or vendors: a security certification from a compliance company is no longer a reliable guarantee that the company has actually secured your data. You now need to audit the auditors themselves and ask detailed questions about what specific vulnerabilities were tested for.