Hallucinations

1 Post

Hallucination Creates Security Holes: Researcher exposes risks in AI-generated code

Language models can generate code that erroneously points to software packages, creating vulnerabilities that attackers can exploit.