Researchers from cybersecurity startup "Hacktron AI" say they gained access to an OpenAI employee’s ChatGPT account. The incident comes as experts warn about the growing ability of AI models to find and exploit vulnerabilities. “If we can’t control these models now, there’s no way we’ll be able to control future models,” Riki Parikh, AI policy director and former Meta employee, told ABC News. Anthropic currently says Claude is helping build its successor, leading 26% of the company’s AI research and development under human supervision. ▻