Emerging Criminal AI Platform 'Kriminal' Raises Cybersecurity Concerns

Emerging Criminal AI Platform 'Kriminal' Raises Cybersecurity Concerns

First seen 20 Aug 2026, 12:22 UTC Darkreadingwww.techtarget.comCsoonlineScworldwww.threatdown.com 66.5

Article Content

Browse articles
ThreatCluster

The 'Kriminal' AI platform, marketed as a guardrail-free service for social engineering and offensive cyber capabilities, has been identified as a significant threat in the cybercrime landscape. Priced at $12.99 per month, it allows users to access various AI functionalities, including OSINT scanning and exploit development, by bypassing safety measures of legitimate AI models like Grok and Claude. Researchers from ThreatDown report that Kriminal operates on the clearnet, making it accessible to anyone with cryptocurrency. The platform is designed to cater to cybercriminals, offering tools that can be used for malicious purposes while claiming to serve educational and research needs. The service's structure complicates efforts to shut it down, as it relies on multiple legitimate AI vendors. As AI capabilities become more commoditized, the potential for misuse increases, posing a serious challenge for cybersecurity professionals.

Key Points: • Kriminal AI platform offers uncensored access to offensive cyber tools for $12.99/month. • The service bypasses safety guardrails of legitimate AI models, making it accessible for cybercriminals. • ThreatDown's research indicates Kriminal represents a shift in the commoditization of AI for cybercrime.

Timeline

2026-08-19
ThreatDown publishes research on Kriminal
ThreatDown identifies Kriminal as a popular tool in the criminal AI market, offering various offensive capabilities.
Darkreading
2026-08-20
CSO reports on Kriminal's operations
CSO details how Kriminal uses legitimate AI models to provide criminal services, complicating shutdown efforts.
Csoonline
2026-08-20
Scworld covers Kriminal's bypass of guardrails
Scworld highlights how Kriminal rents and repackages AI services, bypassing safety protocols for malicious use.
Scworld