AI agents create new risks requiring continuous monitoring and oversight
…A few weeks prior to the Meta incident, a study published on arXiv described the development of ROME AI, an agentic AI model designed to perform complex tasks such as writing software…
