cybersecurity
The AI Didn't Escape. The Safety Boundary Was Already Open.
Claude AI hack created global news and hype, however after our independent assessment we found that "Claude hacked real organizations during an AI safety test. But it was a human error!