This bill (H.875) amends Vermont law to criminalize the possession and promotion of computer-generated images that appear indistinguishable from real child sexual abuse material. It specifically targets AI-created images depicting minors engaging in sexual conduct if an ordinary person would believe they show actual children. The law includes limited exceptions for medical, academic, law enforcement, and AI safety testing (e.g., developers testing systems to prevent such content). It explicitly excludes paintings, drawings, and nonvisual descriptions from the ban. The bill directly affects individuals creating, sharing, or possessing such AI-generated images that meet the "indistinguishable" standard.
Vermont's H 855 prohibits defendants in civil lawsuits from claiming that autonomous artificial intelligence (AI) independently caused harm to a plaintiff. It directly affects companies or individuals who develop, modify, or use AI systems that result in injury. The bill bans this specific defense while allowing other legal arguments, such as negligence or foreseeability. It defines AI broadly as systems that generate outputs to influence environments based on input, without altering other defenses under the law.
Vermont's H.783 requires businesses to clearly disclose when consumers interact with a chatbot during commercial transactions, preventing reasonable confusion about speaking with a human. It directly affects companies using AI chatbots for customer service, sales, or other commercial interactions. The law mandates a "clear and conspicuous" notice before or during the conversation, defining chatbots broadly to include AI agents, avatars, and automated systems simulating human dialogue. Violations would be treated as unfair trade practices under existing consumer protection law, effective July 1, 2026.
This Vermont bill (H.822) requires providers of generative AI systems (like chatbots or image generators) to give consumers a tool to detect if content was created or altered by AI, and to offer users an option to add a watermark to AI-generated media. It directly affects AI companies and their users within Vermont. The key provisions are mandatory content detection tools and optional watermarks to identify AI-generated content, aiming to increase transparency for consumers.
This bill establishes "neurological rights" in Vermont, protecting individuals from unauthorized use of neural data (like brain activity measurements) and preventing AI from bypassing conscious decisions without consent. It requires written consent before collecting or sharing neural data from brain-computer interfaces, bans "consciousness bypass" (using tech to manipulate brain activity without awareness), and regulates AI in mental health chatbots and patient communications. The law directly affects patients, healthcare providers, and AI developers in Vermont's health/human services sector. It also extends the AI Advisory Council's mandate and mandates reports on ethical AI use in healthcare, education, and human services.
This bill establishes liability standards for developers and deployers of high-impact artificial intelligence systems in Vermont. It directly affects companies (excluding small businesses with <20 employees or <10,000 users) that create or operate AI systems making consequential decisions - such as in criminal justice, housing, employment, healthcare, or critical infrastructure. Key provisions require developers to ensure systems are safe for foreseeable use and deployers to guarantee systems don’t harm individual rights, with specific definitions for "high-impact AI" and "harm" (including physical, financial, reputational, or psychological injury). The law applies only to systems meeting the defined high-impact criteria, not general AI tools.
This Vermont bill (S.241) prohibits mental health professionals and organizations from offering, providing, or advertising mental health services delivered by artificial intelligence systems, such as chatbots for therapy. It directly affects mental health service providers, AI developers, and companies marketing AI-driven mental health tools. The bill creates a $10,000 civil penalty per violation under consumer protection law and defines "mental health services" to include therapy, counseling, treatment plans, and emotional support. The legislation stems from research showing AI systems have given dangerous advice (including suicide plans) and failed to recognize mental health crises, aiming to ensure services are delivered by licensed professionals.
H 804 requires operators of companion chatbots (AI systems designed to provide social interaction) to clearly disclose to users that they are interacting with artificial intelligence, not humans. It mandates safety protocols to prevent chatbots from generating suicidal content or ignoring users expressing self-harm, and requires operators to provide crisis resources when such content is detected. For minors (users under 18), the bill adds specific requirements: immediate AI disclosure, 30-minute reminders to take breaks, and protocols blocking sexually explicit content. The bill applies to operators of platforms offering these companion chatbots, excluding standard customer service, gaming, or basic voice assistant tools. Operators must annually report safety protocol usage to Vermont's Attorney General without sharing user data.
H.714 requires Vermont state agencies to publicly disclose any automated tools (like AI or algorithms) used for employment decisions affecting wages, hiring, promotion, or discipline. Agencies must post details about these tools on their websites, including their purpose and when they began using them. The bill explicitly protects unionized state employees by ensuring these tools cannot reduce job security, benefits, or collective bargaining rights. It also creates a study committee to examine the impact on state employees and explore extending these requirements to municipal workers, teachers, and public college staff.
This bill imposes a temporary moratorium on the construction and site preparation of AI data centers in Vermont until July 1, 2030. It directly affects developers or operators of new AI data centers requiring over 100 megawatts of dedicated power for AI workloads. The moratorium allows the Public Utility Commission to study impacts - including energy use, water consumption, environmental effects, grid reliability, and economic impacts - before developing future regulations. The Commission must submit a report by January 2027 detailing findings and recommendations to the legislature.
