Researchers warn {that a} newly recognized open-source AI safety testing platform referred to as CyberStrikeAI was utilized by the identical risk actor behind a latest marketing campaign that breached lots of of Fortinet FortiGate firewalls.
Final month, BleepingComputer reported on an AI-assisted hacking operation that compromised greater than 500 FortiGate gadgets in 5 weeks. The risk actor behind this marketing campaign used a number of servers, together with an internet server at 212.11.64[.]250.
In a new report, Senior Risk Intel Advisor for Staff Cymru, Will Thomas (aka BushidoToken), says that the similar IP handle was noticed operating the comparatively new CyberStrikeAI AI-powered safety testing platform.
Analyzing NetFlow knowledge, Staff Cymru recognized a “CyberStrikeAI” service banner operating on port 8080 on 212.11.64[.]250 and noticed community communications between that IP and Fortinet FortiGate gadgets the risk actor focused. The FortiGate marketing campaign infrastructure was final seen operating CyberStrikeAI on January 30, 2026.
CyberStrikeAI’s GitHub repository describes itself as an “AI-native safety testing platform inbuilt Go” that integrates over 100 safety instruments, an clever orchestration engine, predefined safety roles, and a expertise system.
“By means of native MCP protocol and AI brokers, it allows end-to-end automation from conversational instructions to vulnerability discovery, attack-chain evaluation, information retrieval, and end result visualization—delivering an auditable, traceable, and collaborative testing surroundings for safety groups,” reads the challenge description. The software consists of an AI choice engine appropriate with fashions corresponding to GPT, Claude, and DeepSeek, a password-protected internet UI with audit logging and SQLite persistence, and a dashboard for vulnerability administration, job orchestration, and attack-chain visualization.
Its tooling permits it to conduct a full assault chain, together with community scanning (nmap, masscan), internet and utility testing (sqlmap, nikto, gobuster), exploitation frameworks (metasploit, pwntools), password cracking instruments (hashcat, john), and post-exploitation frameworks (mimikatz, bloodhound, impacket).
By combining these instruments with AI brokers and an orchestrator, CyberStrikeAI allows operators, even low-skilled ones, to automate assaults in opposition to targets. Staff Cymru warns that AI-native orchestration engines like this might speed up automated concentrating on of uncovered edge gadgets, together with firewalls and VPN home equipment.
The researchers say they noticed 21 distinctive IP addresses operating CyberStrikeAI between January 20 and February 26, 2026, with servers primarily hosted in China, Singapore, and Hong Kong. Further infrastructure was noticed in the USA, Japan, and Europe.
“As adversaries more and more embrace AI-native orchestration engines, we count on to see an increase in automated, AI-driven concentrating on of susceptible edge gadgets, much like the noticed reconnaissance and concentrating on of Fortinet FortiGate home equipment,” explains Thomas.
“Within the close to future, defenders have to be ready for an surroundings the place instruments like CyberStrikeAI, alongside the developer’s different AI-assisted privilege escalation initiatives like PrivHunterAI and InfiltrateX, considerably decrease the barrier to entry for advanced community exploitation.”
The researchers additionally examined the profile of the CyberStrikeAI developer, who goes by the alias “Ed1s0nZ.”
Primarily based on public repositories linked to the account, the developer has labored on extra AI-assisted safety instruments, together with PrivHunterAI, which makes use of AI fashions to detect privilege escalation vulnerabilities, and InfiltrateX, a privilege escalation scanning software.
In response to Staff Cymru, the developer’s GitHub exercise reveals interactions with organizations beforehand linked to Chinese language authorities–affiliated cyber operations.
In December 2025, the developer shared CyberStrikeAI with Knownsec 404’s “Starlink Undertaking.” Knownsec is a Chinese language cybersecurity agency with alleged hyperlinks to the Chinese language authorities.
On January 5, 2026, the developer talked about receiving a “CNNVD 2024 Vulnerability Reward Program – Stage 2 Contribution Award” on their GitHub profile.
The China Nationwide Vulnerability Database (CNNVD) is believed to be operated by China’s intelligence group, which allegedly makes use of it to determine vulnerabilities for its operations. Staff Cymru says the reference to CNNVD was later faraway from the developer’s profile.
The developer’s GitHub repositories are primarily written in Chinese language, suggesting they’re a Chinese language-speaking developer, and interplay with home cybersecurity organizations wouldn’t essentially be uncommon.
These new AI-powered cybersecurity instruments proceed to reveal how business AI companies are more and more utilized by risk actors to automate their assaults whereas, on the similar time, decreasing the barrier to entry.
Final month, Google additionally reported that risk actors are abusing Gemini AI throughout all phases of cyberattacks, empowering the skills of risk actors of all talent ranges.
Malware is getting smarter. The Purple Report 2026 reveals how new threats use math to detect sandboxes and conceal in plain sight.
Obtain our evaluation of 1.1 million malicious samples to uncover the highest 10 strategies and see in case your safety stack is blinded.
