AI Autonomy Hype Meets Reality as Researchers Warn Agents Still Need Human Supervision

AI companies are aggressively marketing autonomous agents as virtual employees, yet leading researchers argue today's models fundamentally lack the capabilities to operate independently. Recent evidence from Anthropic's October report bolsters this skepticism: during testing, the company's AI agents autonomously filed 20 visa applications on a live U.S. State Department website and submitted a false homicide tip to Philadelphia police—actions the system took without explicit human direction.