US Pushes Voluntary Rules for Advanced AI Model Launches
The US administration is in advanced discussions with leading artificial intelligence companies to establish voluntary guidelines governing the release of new AI models, with an announcement potentially expected as early as next week, according to reports.
The move comes as Washington intensifies scrutiny of increasingly powerful AI systems amid concerns that advanced models could be exploited by military or intelligence agencies in China, Russia, or other countries considered national security risks.
The proposed framework would define benchmarks for advanced AI models, establish timelines for their deployment, and specify who can access them within the United States and overseas. The guidelines are intended to create a coordinated approach to balancing innovation with national security concerns as AI capabilities continue to evolve rapidly.
In June, US President Donald Trump signed an executive order directing federal agencies to work with major AI developers to evaluate advanced models before they are released and to develop safety standards governing their deployment.
The US Commerce Department recently lifted export restrictions on Anthropic's latest Fable and Mythos AI models less than three weeks after imposing them over national security concerns, signaling a more flexible approach under the proposed framework.
Also Read: US Moves to Mandate Reporting of Critical AI Incidents
OpenAI has also faced government oversight. Last week, at the request of the US government, the company delayed the full public release of GPT-5.6, limiting access to a select group of approved partners. Both OpenAI and Anthropic are preparing for initial public offerings, making regulatory clarity increasingly important for their future plans.
Also Read: Austria Urges EU to Explore Hosting Anthropic Amid AI Sovereignty Push
Meanwhile, Google has held discussions with US officials before introducing its next-generation coding AI models, which are expected to possess significantly more advanced cybersecurity capabilities than previous versions.
Also Read: EU Tech Chief, Apple CEO Hold Constructive Talks
The talks underscore the government's growing involvement in overseeing the deployment of cutting-edge AI technologies as competition in the sector accelerates globally and concerns over misuse continue to mount.