Researchers Identified Uncensored AI Service Luciferus
Sophos analysts uncovered a subscription-based AI model on the Exploit forum designed to bypass ethical restrictions.
Updated on Sept. 26, 2026 in Artificial Intelligence
Live Poll
Is the growth of unregulated AI services making it too easy for individuals to conduct cybercrime?
Sophos X-Ops published research on August 24, 2026, detailing an uncensored AI subscription service called Luciferus. The service is marketed on the Exploit underground forum as an alternative to mainstream AI tools.
Why it matters
The emergence of specialized, uncensored AI models signals a shift in how threat actors may automate cyberattacks. By providing stable access to non-restricted responses, services like Luciferus aim to bypass the safety guardrails common in commercial LLMs.
The service claims to utilize a proprietary model featuring 120 billion parameters. These specifications are marketed to potential users as a high-capacity alternative to jailbroken mainstream platforms.
The players
Sophos X-Ops
This is a specialized cybersecurity research division focused on threat intelligence and analyzing emerging attack vectors.
Exploit
This is a well-known underground forum used by malicious actors for trading, selling, and discussing cyberattack tools.
The details
The Luciferus service operates similarly to traditional malware-as-a-service kits, allowing threat actors to purchase access to uncensored language models. During testing, researchers confirmed that a version dubbed Luciferus Junior successfully generated functional code for a remote access trojan.
Timeline
August 24, 2026: An Exploit forum user advertised the Luciferus service.
The Tech Race
This development follows the established industry pattern of malware-as-a-service ecosystems. By applying these commercial distribution models to large language models, the service replaces the need for unreliable jailbreaks with a stable, paid product.
For users and developers, this indicates that malicious actors have easier access to tools for generating harmful code. Organizations should monitor for an increase in automated phishing and trojan development enabled by these specialized, uncensored models.
The takeaway
The move toward dedicated, uncensored AI tools represents a professionalization of cybercrime that lowers the technical barrier for attackers. Readers should remain vigilant against evolving automated threats that can now leverage high-parameter models for malicious tasks.
Further reading
For broader context on how emerging tools are reshaping the security landscape, visit Artificial Intelligence.
Source note: This article includes information reported by Upgrade Magazine.
Live Poll
Is the growth of unregulated AI services making it too easy for individuals to conduct cybercrime?
