Most states have kicked off their 2026 legislative session, and AI-related bill introductions are rolling in. As of Jan. 29, 28 state legislatures will consider AI-related proposals. The Learning Agency looked into these bills to see which ones could impact students and in what ways.
We saw a few areas of focus:
- Regulating AI chatbots. State legislators are trying to make chatbot companions safer for young people with bills that prohibit chatbots with human-like characteristics, require chatbots to disclose that they are not human, and limit minors’ access to chatbots.
- Protecting children from social media’s harms. New state proposals are attempting to lower the risks of social media for users under the age of 18 by requiring age verification, parental consent, and default privacy settings for minors.
- Increasing transparency about the data used to train AI. A few state bills would require developers to publicly disclose details about their AI training data.
- Curbing AI usage in K-12 schools. While no current proposal represents a full ban on AI in classrooms, a few bills seek to limit students’ access to it.
The majority of student-related bills we examined appear to be in response to public concerns over youth engagement with chatbot companions and social media. Studies, like this one from Common Sense Media and Stanford, and reporting from major outlets (e.g., CBS News, The Guardian, NPR) have raised public awareness of the prevalence and potential dangers of AI companions. Meanwhile, reputable medical sources (e.g., Mayo Clinic, Johns Hopkins Medicine) and researchers like Jonathan Haidt have shed light on social media’s detrimental effects on young people. Policymakers on both sides of the aisle are trying to do something about it.
How State Legislators Are Seeking To Shape AI For Students
Bills with potential educational impacts fall into a few broad categories. In general, state legislators are attempting to mitigate AI’s impact on young people by:
Regulating AI chatbots. Many of the AI-related state bills under consideration aim to restrict access to, or improve the safety of, chatbot companions. Some of the proposals attempt to make AI chatbots less human-like, and others seek to limit minors’ access to them. While the bills listed below are not education-specific, any of them could affect how chatbots are used in classrooms.
Bills to watch:
- In Florida, Senator Leek’s (R) Florida AI Bill of Rights (SB 482) has a provision that requires companion chatbot platforms to prohibit minors from having an account unless their parent/guardian consents. Backed by Governor DeSantis, this bill also creates an “AI Bill of Rights” for Floridians and prohibits AI technology companies from selling or disclosing users’ personally identifiable information
- In Maine, Representative Gramlich (D) introduced LD 2162 with a set of bipartisan cosponsors. This bill prevents access to AI chatbots with human-like features and companion interfaces for minors .
- In Missouri, Representative Schmidt (R) introduced the Children Harmed by AI Technology Act (CHAT), HB 2031. This kids chatbot safety bill requires age verification and parental consent for minors. Oklahoma’s SB 1521, introduced by Senator Hamilton (R), is a similar bill.
- In Virginia, Delegate Maldonado (D) and Delegate Runion (R) introduced the AI Chatbots Act (HB 635) and AI Chatbots and Minors Act (HB 758), respectively. Whereas the former requires a permanent disclaimer to users of all ages that a companion chatbot is not human, the latter prevents chatbots with human-like features from being available to minors.
The majority of student-related bills we examined appear to be in response to public concerns over youth engagement with chatbot companions and social media. Studies and reporting from major outlets have raised public awareness of the prevalence and potential dangers of AI companions. Policymakers on both sides of the aisle are trying to do something about it.
Aiming to protect children from the risks of social media. Other state-level proposals target social media. They promote age verification, increased privacy measures for users under the age of 18, and crack down on the sharing or selling of children’s data. While these are not education-specific bills, they would certainly impact students.
A bill and ballot measure to watch:
- In South Carolina, Representative Guffey (R) introduced the Stop Harm From Addictive Social Media Act, HB 4591. This bill has more than a dozen cosponsors (all Republican) and would make social media platforms verify the age of all users, require parental consent for minors, and establish default privacy settings for minors.
- In California, Common Sense Media filed the Parents and Kids Safe AI Act, a ballot measure backed by OpenAI that would require age verification on social media platforms and ensure child-protective settings are applied for users younger than 18. Additionally, it would prohibit child-targeted advertising and the sale or sharing of minors’ data without parental consent. This ballot measure contains other provisions that are not specific to social media, including several safeguarding requirements for AI systems. If the measure gets about 550,000 voter signatures, it will be on California’s ballot in Nov. 2026.
Calling for transparency around AI training data. Some bills are aimed at shining more sunlight on the data being used to train AI models or services, presumably to promote the usage of high-quality, privacy-protected data. These proposals would make product developers responsible for disclosing the data they feed into their AI models. While they are not education-specific, they would apply to ed tech developers leveraging AI in their products.
Bills to watch:
- In New York, Senator Gounardes (D) and Assemblymember Bores (D) introduced companion versions of the AI Training Data Transparency Act (S 6955 and A 6578, respectively). These bills require developers of generative AI models or services to share on their website information about the datasets used to train the AI model or service.
- In Washington State, Representative Shavers (D) introduced HB 2503 to require developers to disclose on their website details about the data they are using to train their generative AI system.
Many of the AI-related state bills under consideration aim to restrict access to, or improve the safety of, chatbot companions. Some of the proposals attempt to make AI chatbots less human-like, and others seek to limit minors’ access to them.
Curbing AI Usage in Schools. So far, only a couple of bills introduced in state legislatures are specific to AI in school settings. These bills could be characterized as overreaches, as they would ban AI or AI chatbots in classrooms – even AI-powered ed tech with proper safeguards and a record of improving learning outcomes. Neither of the bills highlighted below have cosponsors, an indication that they lack the support needed to pass.
Bills to watch:
- In New York, Assemblymember Carroll (D) introduced NY A09190. This bill would ban the use of AI in classrooms prior to ninth grade, unless used for diagnostic purposes or interventions for students with disabilities. The legislative text makes clear that it would not prevent teachers or other school staff from using AI for administrative or planning purposes. However, it would prevent students in elementary or middle school from engaging with any AI-powered tools or platforms, regardless of their demonstrated effectiveness or safety guardrails.
- In Virginia, Delegate Rasoul (D) introduced HB 1186, which requires school boards to adopt policies to prohibit any student from being “required, encouraged, or permitted to interact with an AI chatbot in order to receive instruction or otherwise complete any lesson or assignment in any course or class.” This bill has broader educational implications than other chatbot bills, as it would not allow academic support chatbots that lack human-like features and disclose that they are AI-powered – even if they have a track record of improving student achievement.
So far, only a couple of bills introduced in state legislatures are specific to AI in school settings. These bills could be characterized as overreaches, as they would ban AI or AI chatbots in classrooms – even AI-powered ed tech with proper safeguards and a record of improving learning outcomes.
Conclusion
It’s still early in state legislative sessions, with some not even open yet, but the AI bills introduced so far show how policymakers are attempting to mitigate AI’s risks for young people. With a focus on chatbot companions, social media platforms, and AI training data, Democrats and Republicans are seeking to curb potential harms and promote transparency, consent, and safeguards.
None of these bills would ban AI outright in schools, but a few attempt to reduce AI usage in classrooms. Those bills represent extremes and lack cosponsors, but they are worth watching. As U.S. students’ learning outcomes stagnate and news stories about the harmful effects of AI companions and social media proliferate, policymakers will seek to respond.
Tasha Hensley
Policy Director