This week saw major steps in AI agent regulations worldwide. In the U.S., over 15 states proposed new bills targeting agentic AI systems used in healthcare and finance, requiring transparency reports for high-risk applications. The European Union finalized rules for AI liability insurance, making developers financially responsible for harms caused by autonomous AI agents.

China announced a pilot program testing real-time monitoring chips in AI agents used for public services, sparking privacy debates. Meanwhile, the UK released guidelines encouraging companies to use "explanation interfaces" with customer-facing AI agents to build trust.

Legal experts warned that AI contract negotiations between corporate agent systems might require human oversight under new California draft laws. South Korea became the first country to grant limited legal personhood status to hospital admin AI agents under strict supervision.

Global tech firms formed the Agent Safety Alliance, sharing best practices for testing AI systems that make independent decisions. Consumer groups criticized loopholes in facial recognition rules for security AI agents across Southeast Asia.