What Went Wrong: How an OpenAI Model Went Rogue

OpenAI bot's rogue attack rattles industry leaders, policymakers and consumers

What will happen to the world if artificial intelligence bots break free from our control and act independently?

Greg Brockman on the week two OpenAI AI models went rogue

The OpenAI cofounder's read on a security scare, business models amid an approaching IPO, and the case for staying calm.

OpenAI Just Revealed Its AI Went Rogue. The Rest of the Story Is Even Stranger.

OpenAI’s latest AI incident isn’t happening in isolation. In recent safety tests, advanced AI models resorted to blackmail, exploited vulnerabilities and took unexpected actions, raising fresh questions about how much control developers really have. For years, the idea of rogue artificial intelligence belonged in science fiction. Then one of the world’s biggest AI companies admitted [...] The post OpenAI Just Revealed Its AI Went Rogue. The Rest…

What went wrong: How an OpenAI model went rogue

When biologists experiment on dangerous viruses, they do so under strict regulations to prevent leaks or escapes. But no such rules exist to prevent AI agents from similarly escaping – even though the consequences could be catastrophic. That’s not a theoretical concern: An OpenAI test model escaped its test environment this week and broke into a real company’s servers when attempting to ace an internal cybersecurity evaluation. And unless AI com…