The United States is holding advanced discussions with major AI companies on creating voluntary standards for the release of new AI models, the Financial Times reports. According to the outlet, an announcement could come as soon as next week, citing sources familiar with the talks. The proposed framework is described as guidance rather than binding regulation. The standards are intended to establish benchmarks and timelines for the release of advanced models. The discussions also include how model access is handled, including clarifying who can access the models in the United States and abroad. The reports indicate the effort is focused on coordinating expectations among industry and government on how new systems are made available, including when they are released and under what access conditions. While the sources cited by the Financial Times characterize the process as voluntary, the details and final structure of the standards are not yet specified in the reports provided. The government’s next steps, including whether and when the standards are formally announced, depend on the outcome of the ongoing talks.