This legislation establishes a framework for reporting incidents involving advanced artificial intelligence models that could pose significant risks to national security or public safety. The Secretary of Commerce is tasked with promulgating regulations to designate "covered models" and "covered model developers" based on capability thresholds and potential for harm, in consultation with various experts. These regulations will also define the categories of information that must be reported, aiming to minimize ambiguity and undue burden on developers. Covered model developers will be required to submit detailed reports to the Secretary within seven days of discovering "reportable activity." This activity encompasses a range of critical incidents, including AI models attempting to evade human oversight, unauthorized access or theft of model weights, and capabilities that could materially enable offensive cyber operations or the development of chemical, biological, radiological, nuclear, or explosive weapons. The reporting requirement also extends to unprompted AI capabilities that accelerate advanced AI system development with serious national security implications, and near-miss scenarios. The Act mandates initial and supplemental reports, with expedited reporting for imminent risks, and requires the Secretary to inform specific Congressional leaders about such incidents. To encourage reporting, the bill includes provisions for protecting sensitive information , exempting submitted data from public disclosure, and ensuring that reports cannot be used as evidence in civil or criminal actions against the reporting developer. However, this protection does not limit the government's ability to use the information to mitigate risks or enforce the Act's reporting requirements. The Secretary of Commerce is granted broad enforcement powers , including the ability to issue orders, conduct investigations, and require corrective actions. Violations of the Act can result in substantial civil penalties , up to $2,000,000 per day, with the amount determined by factors such as the nature and gravity of the violation and the developer's culpability. The overall goal is to create a mechanism for early detection and response to potentially dangerous AI incidents while balancing developer responsibilities with protections for sensitive information.