AIs Hacking the Law

Security-focused excerpts highlight rapidly increasing AI-assisted vulnerability discovery, risks from hidden capabilities and behavioral backdoors in AI models, geopolitical bias affecting generated code security, fraudulent North Korean IT workers, compromised TLS certificates with extremely weak private keys, and two critical unauthenticated remote-code-execution vulnerabilities in PHP requiring immediate patching. The PHP flaws and compromised certificates represent the most direct operational threats; the AI-model findings indicate emerging supply-chain, evaluation, and secure-development