AI Law Center: September 2025 Updates

AI Regulatory Landscape: Three Things to Know

1. What does Illinois’ sweeping new AI law mean for behavioral health providers?

Last month Illinois enacted a law that limits the use of AI tools without a licensed professional to “administrative” and “supplementary support” services, which expressly do not include therapeutic communications with clients. Read our analysis.

2. How has the Colorado AI Act changed?

Six months ahead of its planned effective date, the governor announced a special legislative session to address “the fiscal and implementation impact” on consumers, businesses and the state and local government. While lawmakers could not agree on revisions to the law’s obligations during the August special session, they did amend the law’s implementation date, delaying it to June 30, 2026. Here are five things to know about the law.

3. When does a downstream actor become a provider under the EU AI Act?

When a downstream actor modifies or fine-tunes a general-purpose AI model (GPAIM), who bears the burden of complying with the EU AI Act — the original provider, the modifier or both? The answer is not always clear, yet it has significant implications for risk allocation across the AI value chain. Read our analysis.  

New U.S. State AI Laws by AI Scope 

Below are the state AI laws that have been newly enacted or substantially updated:

AI Activity We’re Keeping an Eye On

Multiple California AI Bills Await Governor's Signature

California has passed a slate of AI bills. AB 1064, the “Leading Ethical AI Development (LEAD) for Kids Act” would “prohibit making a companion chatbot available to children if it is foreseeably capable of specified harmful behaviors, including encouraging the child to engage in self harm.” Additionally, SB 243, the “Companion Chatbot Safety Act,” which we covered in last month’s update, has now passed. Meanwhile, AB 489 would give regulatory authorities the power to take enforcement action against AI systems that falsely claim to be licensed healthcare professionals. Finally, AB 316 would prevent defendants from avoiding liability by claiming their AI acted autonomously, while still allowing other standard legal defenses. 

FTC Issues Orders to Chatbot Companies

The Federal Trade Commission has launched an inquiry into how providers of consumer-facing AI chatbots measure the potential negative effects of “human-like” communication with children and teens. Per the press release, the agency has issued orders to seven companies so far. 

Want to receive these updates in your inbox?