US Government Vets AI Model Releases
- •OpenAI restricts new model release following US government cybersecurity vetting request
- •Anthropic gains administration approval for limited release of its strongest cybersecurity model
- •Both companies limit access to trusted partners and administration-approved customers only
OpenAI and Anthropic are restricting the release of new artificial intelligence models following a vetting process mandated by the administration of US President Donald Trump. This government oversight aims to identify and mitigate potential cybersecurity risks associated with advanced AI products. OpenAI stated it is complying with the request, limiting access to its latest model to a small group of customers specifically approved by the administration.
Anthropic announced that the Trump administration has granted approval for a limited release of its most powerful cybersecurity model. This development occurs just two weeks after the US Commerce Department had effectively banned the model. According to the companies, these restricted releases are currently available only to select groups of trusted partners and government-approved entities. These measures mark an unprecedented level of federal government involvement in the deployment of AI systems. The companies confirmed that broader public availability for these specific models remains under the purview of federal review processes as they navigate the ongoing security assessment requirements.