Governments around the world are creating new rules for artificial intelligence, especially for the powerful new agentic AI systems that can work on their own. The European Union published new guidelines this week about how companies must be honest when they use AI to create content. The United States is working to create one national rule instead of many different state rules. Meanwhile, companies building AI agents are working together to make sure different AI systems can talk to each other and follow the same rules.

The problem right now is that AI is growing so fast that rules cannot keep up. Many companies say they do not have good systems to check if their AI is working correctly and fairly. The good news is that banks and financial companies think regulators will start helping companies use AI more, instead of just blocking it. Experts say that in 2026, the big challenge will not be building AI agents, but making sure they can be trusted and controlled when thousands of them work together.