Bitcoin researcher Rob Hamilton lost access to OpenAI tools and turned to Chinese AI models to continue his security work. His team used open-source AI systems like Kimi K3 and GLM 5.2 to audit Bitcoin code and find vulnerabilities. You may be wondering how this shift affects cybersecurity and AI access.
Why the Switch Happened
Hamilton, CEO of Bitcoin custody firm AnchorWatch, lost access to OpenAI’s trust cyber program without explanation. He had completed all required steps, but his access was revoked. This forced his team to find alternative AI tools. You might be thinking, what does this mean for security research?
The audit involved 390+ open-source Bitcoin repositories and uncovered thousands of issues. The project cost between $20,000 and $40,000. Hamilton’s team used a mix of AI systems, including Kimi K3 and even OpenAI’s own Cyber Harness. This shows that the audit wasn’t abandoned — it was just rerouted.
What This Means for AI and Security
This situation highlights a growing divide between AI developers and security researchers. Many in the crypto industry are waiting for access to powerful AI tools. Only a few have been granted that access, creating a fragmented landscape.
OpenAI and other U.S.-based companies have strict access policies. Chinese AI models offer an alternative, but they come with their own challenges. You may be wondering how this affects the future of AI and security.
Practitioners in the field are paying attention. Many see this as a symptom of a larger issue: the tension between AI innovation and security. As AI becomes more central to cybersecurity, who gets to use it and under what conditions will shape the future of digital defenses.
The Broader Implications
The barriers to AI access are creating a complex environment. Security researchers must navigate strict policies while attackers find new ways to exploit systems. You may be thinking, what can be done to make AI more accessible and secure?
This shift shows that AI tools are not just for developers. They are essential for security professionals who need to find and fix vulnerabilities. As AI models become more powerful, the debate over who controls them will only grow louder.
For those on the front lines of cybersecurity, the stakes have never been higher. You may be wondering what the next steps are and how you can stay informed about AI and security developments.
