Your AI Coding Agent Is the Attack Surface Now
A fake bug report hijacked coding agents at a Fortune 100 company. OWASP says prompt injection may be permanent. Three new pieces of research just redrew the security map for anyone running AI agents in production.