A decision made in Washington briefly left British users without access to one of the world’s most advanced AI models. Now, the UK government is examining what that kind of disruption could mean for its economy and national security, raising a broader question about how much control governments actually have over the AI systems their businesses and public institutions increasingly rely on.
The review was triggered by the U.S. government’s June decision to restrict foreign access to Anthropic’s Fable 5 and Mythos 5. And the Cabinet Office has asked the Department for Business, Innovation, Science and Trade (DBIST) to assess the economic and security consequences of losing access to frontier AI models from companies such as Anthropic and OpenAI.
This episode has given the UK a real example of what AI dependency can look like when access to a critical service is controlled outside the country.
How The 19-day Disruption Happened
Anthropic launched Fable 5 and Mythos 5 on June 9. Three days later, the U.S. government issued an export control directive requiring Anthropic to suspend access for foreign nationals, including those outside the United States.
Anthropic said it could not reliably determine users’ nationality in real time, so it disabled access to both models for everyone. The U.S.’s reason behind this decision was on national security grounds and was linked to concerns about a method of bypassing Fable 5’s safety protections.
While the restrictions were lifted on June 30, with Fable 5 returning globally on July 1, the gap still created a 19-calendar-day period in which the model was unavailable, including to users in the UK.
The disruption was particularly notable because Fable 5 was not a niche research system, as Anthropic only positioned it as a frontier model for software engineering, professional work, scientific research, and other complex tasks.
Why the UK is Now Looking More Closely
The British government had already recognised that advanced AI is becoming strategically important. Its AI Opportunities Action Plan calls for greater UK influence over frontier AI and supports investment in domestic companies, computing infrastructure, and access to strategically important technologies. The government has also backed a Sovereign AI Unit with up to £500 million, serving as a direct investment in domestic artificial intelligence startups in a bid to keep technology capabilities on British soil.
But the Fable 5 episode exposed a different problem. Building more domestic computing capacity does not automatically guarantee access to the most capable models. A country can have data centres, chips, and AI researchers while still depending on foreign companies for the models running on that infrastructure.
And this concern was also raised in Parliament during the June outage.When asked during Commons oral questions, Foreign Office Minister Seema Malhotra said the withdrawal of the models “reinforces the importance of Britain having greater sovereign control over critical technology and maintaining our national capabilities for us and our allies.”
The Bigger Question For Governments
The UK review is, therefore, looking beyond one temporary outage. Agencies are assessing what could happen to economic growth and cybersecurity if British organisations were denied timely access to the latest AI models. And this is so because frontier AI is already being used to automate work across areas including software, legal services, and accounting, according to people familiar with the government’s assessment.
For other countries, the lesson from the Fable 5 shutdown is that access to advanced AI should not depend on decisions made by companies and governments beyond their borders.
The UK is now trying to work out how much dependence is acceptable and what capabilities it needs to control for itself. And its AI Security Institute, which sits within the DBIST, may be in charge of this initiative.
