This bill, known as the "FRONTIER Act," aims to establish a comprehensive federal oversight system for the development and deployment of frontier artificial intelligence (AI) models in interstate and foreign commerce. Its primary goal is to mitigate potential catastrophic risks associated with advanced AI, defining such risks as those foreseeably contributing to the death or serious injury of over 50 people or more than $1 billion in property damage from a single incident. The legislation introduces the role of an Under Secretary of Commerce for AI Security, responsible for prescribing regulations and overseeing the new framework. The bill categorizes AI developers based on their scale, imposing different obligations. A large frontier developer is defined by gross revenues exceeding $50 million and AI-related development expenditures of at least $1 billion over 36 months. A very large frontier developer has even higher thresholds, with gross revenues over $5 billion and AI-related development expenditures of at least $10 billion. These definitions ensure that the most impactful AI entities are subject to the strictest oversight. Large frontier developers are mandated to create, implement, and publicly publish a "frontier AI framework" detailing their protocols for managing, assessing, and mitigating catastrophic risks. This framework must incorporate widely accepted risk mitigation practices, identify risk thresholds, and outline procedures for reviewing assessment results before model deployment. Annually, these developers must also retain an independent third party to audit their compliance with this framework, with audit findings and summaries published publicly. Furthermore, frontier developers must publish transparency reports before deploying new or substantially modified models, detailing aspects like intended utilization, supported languages, and catastrophic risk assessments. The bill also establishes mechanisms for confidential reporting of critical safety incidents and catastrophic risks to the Under Secretary, with strict timelines for reporting. Failure to comply with these transparency and reporting requirements can result in significant civil penalties. For very large frontier developers, the bill introduces a requirement to retain Independent Verification Organizations (IVOs) , licensed and overseen by the Under Secretary. These IVOs will conduct ongoing assessments of the developer's AI framework, governance, risk monitoring, and mitigation strategies to ensure acceptable levels of catastrophic risk mitigation. IVOs must submit regular assessment reports, including any identified deficiencies and recommended corrective actions, to both the developer and the Under Secretary. The legislation includes stringent conflict-of-interest rules for IVOs to ensure their independence from the AI industry. IVOs are granted immunity from suit and liability for catastrophic risks, except in cases of willful misconduct. The Under Secretary also has the authority to require ad hoc assessments and IVOs must submit supplemental reports if prior findings of acceptable risk mitigation are no longer valid or if an imminent catastrophic risk is identified. A critical provision grants the Secretary of Commerce the authority to issue emergency orders to suspend or restrict a frontier developer's AI activities if an "imminent catastrophic risk" is found. These orders can be provisional or final, with specific content requirements, timelines, and judicial review processes. Willful violation of an emergency order carries severe civil and criminal penalties, including fines up to $10 million per violation and up to 10 years imprisonment. Finally, the bill includes preemption provisions, stating that no state or political subdivision may adopt or enforce laws imposing new substantive obligations on AI developers regarding frontier AI risk transparency, third-party auditing, or incident reporting. However, it explicitly preserves state authority to enforce generally applicable laws, regulate AI use by deployers or users, protect minors from AI-related harms, and govern AI procurement by state governments.