This week brought important updates on AI agent regulations worldwide. In Europe, new rules under the EU AI Act are focusing on autonomous AI systems that make decisions without human help. Companies must now check if their AI tools fall under "high-risk" categories, which could mean extra safety checks and paperwork.

American lawyers are debating how AI legal assistants should be managed. Some worry these tools might give bad advice if not properly supervised. California proposed a law requiring clear labels when AI agents interact with people, similar to rules for phone sales calls.

Big tech companies announced a new AI safety partnership to create standard tests for agentic AI. These tests will check if AI can handle unexpected situations safely. Meanwhile, privacy experts warn that AI data collection laws need updating as agents can now gather information from multiple sources automatically.

In healthcare news, Canada introduced special approval process for medical AI agents that help diagnose patients. Doctors will need to double-check all AI recommendations until the system proves reliable over time. South Korea became the first country to set rules for AI financial advisors, requiring human oversight for large investments.