This weekly update brings important news about AI agents and how they are being governed around the world. The biggest announcement came from Singapore, which released a new set of rules called the Model AI Governance Framework for Agentic AI on January 22, 2026, at a major business conference in Davos, Switzerland. AI agents are computer programs that can make decisions on their own and take actions without being told exactly what to do each time. Singapore's framework helps companies understand how to keep these powerful tools safe and under control.

In the United States, things are changing quickly for AI companies. On January 9, 2026, the Department of Justice announced a new team called the AI Litigation Task Force to challenge state laws about AI that the government thinks are too strict. Meanwhile, many states passed their own AI rules that started on January 1, 2026, including California's new law that requires AI companies to report safety problems within 15 days and face fines up to one million dollars if they break the rules. Texas also passed laws banning AI systems used to hurt people or break laws, with fines ranging from ten thousand to two hundred thousand dollars. However, these state-by-state rules are complicated, and the government wants to create simpler national rules instead.

The new Singapore framework focuses on four important ideas: limiting what actions AI agents can take, making sure humans stay in charge and responsible, using technology to keep agents under control, and teaching people how to use AI agents safely. Companies deploying AI agents need to test them before using them in real situations and keep watching them carefully even after they start working. The framework is voluntary, meaning companies don't have to follow it, but many are paying attention to Singapore's guidance because it came from government experts who understand AI technology well.

Overall, this weekly update shows that AI governance is becoming serious business globally. Governments, experts, and companies are all working to figure out how to let AI agents help us while keeping them from causing problems.