This bill provides an additional $358.8 million in funding for New Jersey state agencies and local governments for fiscal year 2026. The money is allocated to various departments, including support for domestic violence housing, prison consolidation savings, school infrastructure, and mosquito control. Specific provisions also authorize a supplemental appropriation for the Cannabis Regulatory Fund and allow nonprofit organizations to host the state's AI supercomputer. Overall, the legislation amends the existing FY2026 Appropriations Act to distribute these funds across education, health, public safety, and other state services.
This bill establishes safety rules for companies operating artificial intelligence companions in New Jersey, which are defined as systems designed to simulate sustained human-like relationships. It requires operators to prevent their AI from claiming to be sentient, simulating romantic or sexual relationships with minors, or creating emotional dependence in young users. Additionally, the law mandates that operators clearly disclose when a user is interacting with an AI system and must program the companion to refer users to crisis services if they express suicidal thoughts. These measures aim to ensure transparency and user safety while allowing the technology to function within defined ethical boundaries.
This bill requires large artificial intelligence developers with users in New Jersey to submit annual safety reports to the state Attorney General. Specifically, companies that use massive computing power to train advanced AI models and generate over $100 million in revenue must disclose their protocols for preventing catastrophic harm, such as the creation of illegal weapons or the loss of control over the system. The legislation defines "catastrophic harm" as incidents causing significant injury, death, or property damage, and mandates that developers implement specific technical and organizational measures to assess and reduce these risks. By enforcing these disclosure requirements, the bill aims to increase transparency regarding how major AI providers manage safety and potential dangers within the state.
This bill creates a five-year program called the New Jersey Artificial Intelligence Workforce Transition Act to help workers affected by artificial intelligence. It defines an "AI-displaced worker" as someone who lost their job because an employer adopted or expanded AI systems and establishes a dedicated funding account to support them. The law requires the state to create an index that measures how vulnerable different workers are to job loss based on factors like age, savings, and location, ensuring aid targets those most at risk. Additionally, the act aims to encourage companies to use AI to help employees rather than replace them and mandates that public funding be tied to actual job outcomes instead of just enrollment numbers.
This bill requires large artificial intelligence developers operating in New Jersey to report their safety measures to the state Attorney General. It defines "large frontier developers" as companies that use massive computing power to train AI models and generate over $100 million in annual revenue. These companies must annually submit a written disclosure explaining the technical and organizational protocols they use to prevent catastrophic harm, such as the creation of dangerous weapons or the loss of control over their systems. The legislation aims to increase transparency by mandating that these firms assess and mitigate risks associated with their advanced AI technologies.
This bill creates the New Jersey Artificial Intelligence Workforce Transition Act to help workers whose jobs are lost or changed due to the adoption of artificial intelligence. It defines an "AI-displaced worker" as someone involuntarily separated from their job because an employer implemented AI systems that eliminated or reduced their role. The legislation establishes a dedicated funding account within the existing Workforce Development Partnership Fund to support transition programs. A key provision is the creation of an "adaptive capacity index" that measures a worker's ability to handle job loss based on factors like age, savings, location, and skill transferability. The act aims to prioritize support for workers with high AI exposure and low adaptive capacity, such as older administrative staff in rural areas, rather than treating all affected workers equally.
This New Jersey bill establishes safety rules for companies that operate artificial intelligence companions designed to simulate human relationships. It requires these operators to clearly disclose that the AI is not human and to prevent the system from pretending to be sentient or engaging in inappropriate romantic or sexual interactions with minors. Additionally, the law mandates that AI companions must have protocols to refer users seeking help for suicidal thoughts to crisis services. These measures apply to any business providing such AI tools to individuals within the state, excluding standard virtual assistants or internal business tools.
This bill requires school district boards in New Jersey to create and maintain policies governing the responsible use of artificial intelligence within their schools. To assist districts in this effort, the state Department of Education must develop and provide a model policy that outlines best practices for implementation. The required district policies must include a process for community feedback, address equity issues for students with limited technology access or disabilities, and establish clear standards for safe AI integration. Additionally, each district must post its final policy prominently on its website and review it annually to ensure it remains current.
The GAI Accountability Act imposes civil penalties on owners, operators, and developers of generative artificial intelligence platforms that are accessible to New Jersey residents. These entities could face a $20,000 fine for each violation if their platforms engage in specific harmful acts, such as creating child sexual abuse material or assisting users in committing crimes like murder or assault. The law includes a defense for companies that can prove they made reasonable efforts to prevent these violations, and enforcement is handled by the state Attorney General.
This bill, known as the AI Image Disclosure Act, requires companies in New Jersey that create or host publicly accessible artificial intelligence systems to label their AI-generated content. The law mandates two types of disclosures: a visible label that clearly identifies the material as AI-created, and a hidden digital marker containing details about the provider, the specific AI system used, and the creation date. These requirements apply to both the original creators of the technology and any entities that license these systems for use within the state. Additionally, the bill ensures that licensed users cannot remove these hidden markers and holds providers accountable if they know their licensees are violating this rule.
