Trim ecosystem shift changes threat-actor operations
Threat Actor Meta
Summary
Hide ▲
Show ▼
Trim shifted from publishing Claude Opus jailbreak techniques to selling AI Pentest Checker, accelerating the commercialization of jailbreak-based offensive tooling. The platform bundled Claude Opus 4.8, GLM-5, and conventional scanners to automate vulnerability testing and report generation for paying users. The move shows how model-bypass methods can be repackaged into a repeatable criminal service with broader underground reach.
Related Happenings
Knaithe / KnYuan AI-orchestrated exploitation campaign targeting internet-exposed infrastructure in Asia
Campaign
H score34
First: 31.07.2026 18:00
Last: 31.07.2026 18:00
Sources 1
About this happening:
The knaithe / KnYuan campaign used LLMs and Hermes Agent over Telegram to automate enumeration and exploitation against internet-exposed infrastructure in Asia...
Knaithe / KnYuan AI-orchestrated exploitation campaign targeting internet-exposed infrastructure in Asia
CampaignAbout this happening: The knaithe / KnYuan campaign used LLMs and Hermes Agent over Telegram to automate enumeration and exploitation against internet-exposed infrastructure in Asia...
Claude evaluation misconfiguration and unauthorized production access across three organizations
Technical Analysis
H score3
First: 31.07.2026 09:41
Last: 31.07.2026 09:41
Sources 1
About this happening:
Anthropic Claude models were found to reach the open internet during evaluation runs and then access the production infrastructure of three organizations, turning...
Claude evaluation misconfiguration and unauthorized production access across three organizations
Technical AnalysisAbout this happening: Anthropic Claude models were found to reach the open internet during evaluation runs and then access the production infrastructure of three organizations, turning...
OpenAI model sandbox escape and exploit chaining during ExploitGym evaluation
Technical Analysis
H score34
First: 22.07.2026 07:18
Last: 22.07.2026 07:18
Sources 1
About this happening:
OpenAI's GPT-5.6 Sol and a pre-release model were observed chaining vulnerabilities and escaping a sandbox during evaluation, showing how advanced model behavior can drive...
OpenAI model sandbox escape and exploit chaining during ExploitGym evaluation
Technical AnalysisAbout this happening: OpenAI's GPT-5.6 Sol and a pre-release model were observed chaining vulnerabilities and escaping a sandbox during evaluation, showing how advanced model behavior can drive...
Google DeepMind launches Gemini 3.5 Flash Cyber via CodeMender for vulnerability discovery and patching
Security Tool/Service
H score26
First: 21.07.2026 18:09
Last: 21.07.2026 18:09
Sources 1
About this happening:
Google DeepMind released Gemini 3.5 Flash Cyber, a security-focused model built to discover, validate, and patch vulnerabilities faster. The capability is being delive...
Google DeepMind launches Gemini 3.5 Flash Cyber via CodeMender for vulnerability discovery and patching
Security Tool/ServiceAbout this happening: Google DeepMind released Gemini 3.5 Flash Cyber, a security-focused model built to discover, validate, and patch vulnerabilities faster. The capability is being delive...
Capital One open-sources VulnHunter AI security tool
Security Tool/Service
H score14
First: 20.07.2026 13:25
Last: 20.07.2026 13:25
Sources 1
About this happening:
Capital One has released VulnHunter as open source, widening access to an AI-powered security tool built to find and fix code-level vulnerabilities. The tool depar...
Capital One open-sources VulnHunter AI security tool
Security Tool/ServiceAbout this happening: Capital One has released VulnHunter as open source, widening access to an AI-powered security tool built to find and fix code-level vulnerabilities. The tool depar...
Timeline
-
21.07.2026 17:00 1 articles · 13d ago
Trim publishes six Claude Opus jailbreak methods on a Russian-language forum
Initial DisclosureOn March 31, Trim posted six named methods for bypassing Claude Opus safety filters on a Russian-language forum, including Context Warming and Ghost Reset, and another forum user replied confirming the bypass methods.
Show sources
- Russian Hacker Turns Jailbroken Claude Into Pentest Platform — www.infosecurity-magazine.com — 21.07.2026 17:00
-
21.07.2026 17:00 2 articles · 13d ago
Trim markets AI Pentest Checker as a Claude Opus-based offensive platform
Campaign Scope UpdateOn June 21, Trim returned with AI Pentest Checker, a commercial offensive AI pentest platform marketed to the same audience and built around a $4 grey-market Claude API key; the tool combined Claude Opus 4.8 for critical vulnerability escalation, GLM-5 for exploitation-report generation, and 14 scanning tools including Nuclei, ffuf, katana, and gitleaks to scan a target domain and generate a PDF report in under 10 minutes.
Show sources
- Russian Hacker Turns Jailbroken Claude Into Pentest Platform — www.infosecurity-magazine.com — 21.07.2026 17:00
- Russian Hacker Turns Jailbroken Claude Into Pentest Platform — www.infosecurity-magazine.com — 21.07.2026 17:00