US asks OpenAI and Anthropic to withhold new AI models from the UK

Anthropic / Wikimedia Commons (Public domain) The US government has reportedly requested OpenAI and Anthropic to delay the release of their new AI models to the UK, according to a social media post. This request, reportedly from the White House’s Office of the National Cyber Director, aims to ensure these models undergo US government review before being accessible to UK entities, particularly the AI Security Institute (AISI). This development is part of a broader US strategy to control the deployment of advanced AI technologies. The move may impact the UK’s access to cutting-edge AI advancements, potentially affecting the competitive positioning of companies like Anthropic. This directive appears to align with previous measures by the US to regulate foreign access to advanced AI models. Earlier actions included restrictions on Anthropic’s top models and controlled rollouts by OpenAI. As the UK’s AISI is a major public model-testing body, this withholding could limit its ability to test frontier AI models, impacting its insights into technological advancements and security assessments. Market participants may interpret these developments as potentially affecting Anthropic’s competit...