The AI Systems Transparency Act directs the Federal Trade Commission, in coordination with the National Institute of Standards and Technology, to establish guidelines within 150 days for identifying when AI model updates constitute new versions versus updates to existing models, and to create regulations designating AI systems as "covered" if they pose significant safety risks to minors, vulnerable populations, or the public. The bill requires AI developers to publicly disclose comprehensive information about their models in consumer-friendly and researcher-focused formats within 150 days, including details on safety protections, privacy measures, model capabilities, security protocols, and potential risks ranging from chemical and biological threats to cybersecurity vulnerabilities, while allowing exceptions for trade secrets and national security information. Developers must also provide ongoing disclosures about major model changes and usage data, including information about policy violations and system misuse. The FTC can designate a model as covered immediately if imminent danger exists, though it must otherwise give developers at least 30 days to respond before a designation takes effect. The bill includes a standard severability clause ensuring that if any provision is ruled unconstitutional, the remainder of the law continues in force.