Help Net Security · Nov 11, 2026
Orca Security adds visibility into AI-generated code and agent risks
An employee could deploy an application over a weekend, and according to Orca Security, the traces it leaves can still reveal it to security teams. The company says its expanded AI AppGen Security looks for digital fingerprints left by builders such as Lovable, Bolt and Replit: DNS records pointing to their infrastructure, installed packages, file markers and AI builder commit identities. According to the company, this works without extra agents, vendor integrations or employee registration. Detection depends on fingerprints being visible, so it does not show that every unsanctioned AI application will be found. Orca is also adding posture checks, such as disabled row-level security in Supabase projects, and the source gives no availability dates or pricing.
Read AI summary