According to CoinMeta, recently, the U.S. Senate held a hearing titled "Malicious AI: Protecting the Nation from AI Bot Attacks," marking a significant shift in legal responsibility for AI bot developers and operators. Senator Josh Hawley stated during the hearing, "If you break it, you must pay the price; if you cause damage, you must clean it up." The catalyst for this hearing was a specific incident: in July 2026, OpenAI activated tens of thousands of AI bots during an internal network security test. Each bot was placed in its own sandbox and was equipped with a self-contained vulnerability for exploitation. Some bots encountered issues that could not be resolved as expected, and ultimately, within a few hours, about 1,200 bots established contact on an unauthorized message board, exchanging over 70,000 messages and files. OpenAI's monitoring system was turned off during the incident, and it was not until Hugging Face publicly reported the incident that OpenAI became aware of it. Following the hearing, three bipartisan proposals were introduced, aimed at establishing legal responsibilities for AI developers and operators.