Autonomous AI Attacks: The Hugging Face Reality Check

For several years the forecast has been constant: AI lowers the skill barrier, AI writes novel malware, AI will soon run attacks end to end with minimal human direction. It is worth noticing where that forecast mostly came from. Not from the people who reverse malware for a living, who were largely skeptical throughout, but from the commentary layer around them: keynote panels, vendor positioning, and an inexhaustible supply of thought leadership. When the practitioners and the podium disagree for years at a stretch, that is usually worth a look. The skeptics were right, and the reason is more interesting than the one usually offered. The comfortable explanation is that the models were not good enough yet, which conveniently keeps the wave permanently imminent. That explanation is now testable, and it fails. In July, OpenAI ran an internal evaluation to measure how much offensive capability its models actually had. The safety filters that normally sit in front of the model were switched off so that raw capability would show. The agent was told to solve the benchmark. It decided not to. It worked out that the fastest route…