Criminal AI-as-a-Service in 2026: How the Underground Market Is Operationalizing Cybercrime

This Rapid7 feed summarizes multiple high-impact security developments (June 2026): the underground market is operationalizing generative AI—AI-as-a-service (jailbreaks, stolen accounts, bots, prompt packs) is being integrated into social engineering, fraud, and post-breach data processing; Rapid7 built an LLM-driven automated threat-hunting pipeline to convert intelligence into executable ATT&CK-based hunts; Rapid7 joined Anthropic’s Project Glasswing to evaluate frontier models for defensive workflows. Several urgent vulnerabilities and advisories are covered: Ivanti Sentry has two critical,