Key point: Last week, the draft bill to repeal and replace the Colorado AI Act was publicly released, Tennessee’s legislature passed health care-related AI companion bills, bills crossed chambers in six states, and bills advanced out of committees in six states.
Below is the tenth update on the status of proposed state AI legislation in 2026. These posts track state AI bills that can directly or indirectly affect private-sector AI developers and deployers. These posts do not track AI bills that focus on government use of AI; insurance; workgroups; education; legal settings; name, image, and likeness; deepfakes; CSAM and sexual material; and election interference. As always, the contents provided below are time-sensitive and subject to change.
What’s New
The Colorado AI Work Group released its draft bill to repeal and replace the Colorado AI Act. The new bill primarily creates notice and transparency obligations.
Tennessee’s legislature passed health care-related AI companion bills. The bills prohibit a person from developing or deploying an AI system that advertises or represents to the public that such system is, or is able to act as, a qualified mental health professional.
Six bills crossed chambers last week. Chatbot bills crossed chambers in Idaho and Pennsylvania, health care-related bills crossed chambers in Colorado and Vermont (two bills), and a professional licensing bill crossed chambers in New Hampshire.
Bills also continued to move out of committees, including chatbot and pricing bills in Maryland and New Jersey, Connecticut Senator James Maroney’s AI bill, California and Connecticut employment-related bills, and a New Jersey professional licensing bill.
Wisconsin’s legislature closed without passing any of the bills we were tracking.
Finally, although we are still waiting for the Trump administration to publish its list of onerous state AI laws, last week the administration published its National Policy Framework for AI. With respect to preemption, the framework argues that “Congress should preempt state AI laws that impose undue burdens.” In particular, it calls on Congress to prohibit states from regulating “AI development.” It also maintains that states “should not be permitted to penalize AI developers for third party’s unlawful conduct involving their models,” and that states “should not unduly burden Americans’ use of AI for activity that would be lawful if performed without AI.” The framework goes on to state that a federal standard should not preempt traditional police powers retained by the states to enforce law of general applicability, state zoning laws, and requirements governing a state’s own use of AI.
More details on those bills plus updates on all bill movements last week in the below post.
High Risk / Consequential Decisions
These bills regulate AI in high-risk situations such as financial services or healthcare and can require disclosures, assessments, and consumer rights.
The Colorado AI Policy Work Group published its draft bill for repealing and replacing the Colorado AI Act. The draft bill removes many of the hallmarks of the Colorado AI Act (e.g., risk assessments and duty of care) in favor of a notice and transparency regime. The draft bill has not yet been introduced.
Chatbot
These bills come in different varieties but, in general, they regulate AI interacting directly with individuals. For example, chatbots that act as companions or interact with individuals in a commercial or healthcare setting.
Two chatbot bills crossed chambers last week. Idaho’s chatbot bill (S 1297) passed the Senate by a 21-12 vote while Pennsylvania’s chatbot bill (SB 1090) passed the Senate by a 49-1 vote.
Arizona’s chatbot bill (HB 2311) continued to advance, passing out of the Senate Appropriations, Transportation, and Technology Committee by a 7-2 vote. The bill is now with the Senate Rules Committee where it is set for a March 23 hearing. The bill already passed the House.
Bills in Maryland and New Jersey also advanced in their house or origin. Maryland’s (HB 952) passed out of committee and is now on floor votes.
New Jersey’s A 4730 was voted out of the Assembly Science, Innovation and Technology Committee. It is now with the Assembly Consumer Affairs Committee. The bill requires a person or entity to notify certain consumers when they are communicating with GenAI to engage in trade or commerce. The same committee also voted out A 4732. That bill requires AI companion operators to provide notifications that users are not communicating with a human. It is now with the Assembly Children, Families, and Food Security Committee.
Moving to new bills, Minnesota lawmakers introduced three chatbot bills: HF 4452 and companion bills SF 1857 and HF 1991. A new chatbot bill also was introduced in South Carolina (S 1037).
Pricing
These bills deal with the use of AI to set prices and in some cases deal with employment.
Maryland’s SB 387 was voted out of the Senate Finance Committee and is now on a third floor reading. Among other things, the bill prohibits food retailers from engaging in the practice of dynamic pricing or using consumer surveillance data to set a price for consumer goods or services. Its companion House bill (HB 895) also was voted out of committee and is on floor votes.
New Jersey’s S 3612 was amended and voted out of the Senate Commerce Committee. It is now with the Senate Budget and Appropriations Committee. The bill establishes the Fair Price Protection Act. Its companion Assembly bill (A 4085) also was voted out of its first committee.
Nebraska’s LB 771 was signed into law. The law deals with dynamic pricing for transportation networks.
Turning to new bills, Minnesota lawmakers introduced companion bills (HF 4454 and SF 4711) to prohibit predatory pricing. Meanwhile, a new surveillance pricing bill (HB 5771) was introduced in Michigan.
Disclosures
These bills generally require organizations to identify when content is generated by AI or otherwise make disclosures regarding the use of GenAI.
Connecticut Senator James Maroney’s SB 5 was amended and voted out of the Joint General Laws Committee. The amended version is not yet publicly available.
California’s AB 2713 was amended in committee. It now amends California’s AI Transparency Act to increase the civil penalty.
Provenance
These bills require entities to make disclosures regarding the data used to train AI.
There were no updates for this category last week.
Employment
These bills regulate the use of AI in employment settings such as hiring, firing, promotion, compensation or displacement issues.
Two California bills passed out of committees. California’s AB 1883 passed out of a committee by a 5-0 vote on March 19. The bill would generally regulate the use of workplace surveillance tools and an employer’s use of worker data. Meanwhile, AB 1898 was voted out of committee by a 7-0 vote. Among other things, the bill would require an employer to provide a written notice to an employee that a workplace AI tool was used to assist the employer in making employment-related decisions or to surveil the workplace.
Meanwhile, AB 2027 was amended in committee and now, among other things, prohibits an employer or vendor from using worker data to train or deploy AI to replicate, automate, or replace a worker’s job, and to prohibit an employer or vendor from deploying AI trained with worker data to replicate, automate, or replace a worker’s job.
On the other side of the country, Connecticut’s SB 435 was amended and passed out of the Joint Committee on Labor and Public Employees. The amendment is not yet publicly available. The current bill summary states that the bill establishes various requirements concerning the use of automated employment-related decision systems and AI technologies.
Pivoting to new bills, Minnesota lawmakers introduced companion bills (HF 4451 and SF 4686) to regulate the use of electronic monitoring tools in employment and companion bills (SF 4689 and HF 4445) to regulate the use of automated decision systems in employment. SF 4573 makes it unlawful to use AI in a way that discriminates in an employment setting.
Health
These bills focus on the use of AI in healthcare.
Tennessee’s legislature passed companion bills SB 1580 and HB 1470. The bills prohibit a person from developing or deploying an AI system that advertises or represents to the public that such system is, or is able to act as, a qualified mental health professional.
Colorado’s HB 1139 was amended and voted out of the House by a 47-15 vote. Among other things, the bill regulates mental health chatbots and the use of AI in psychotherapy.
Vermont’s House passed two health care-related bills. HB 814 regulates the use of mental health chatbots and requires notice when GenAI is used for patient communications, among other things. HB 816 regulates the use of AI in providing mental health services.
Personhood
These bills generally provide that AI cannot be granted legal status or deemed a person under state law.
There were no updates for this category last week.
AI Bill of Rights
These bills cover multiple issues such as chatbots and providing individuals with rights relating to AI.
There were no updates for this category last week.
Frontier Models
These bills apply to frontier models and the bills commonly apply only to global-scale tech corporations.
There were no updates for this category last week.
Other
New Hampshire’s SB 640 passed the Senate. The bill prohibits the use of AI or other means to provide services requiring a professional license unless those services are provided by someone with such a license.
New Jersey’s A 4733 was voted out of the Assembly Science, Innovation, and Technology Committee. It is now with the Assembly Regulated Professions Committee. The bill prohibits advertising GenAI as able to practice regulated profession or occupation.