Restricted Access to Anthropic’s Mythos AI Sparks Calls for Stronger UK AI Sovereignty in Banking

The limited ability of UK banks to access Anthropic's advanced artificial intelligence model, Mythos, has intensified concerns over Britain's dependence on foreign technology and highlighted the need for the country to accelerate the development of its own AI capabilities. Harriet Rees, Chief Information Officer at Starling Bank and the government's appointed "AI champion" for the banking sector, said the situation demonstrates why the UK must take a long-term strategic approach by investing in domestic AI infrastructure, developing homegrown AI models, and strengthening national expertise rather than relying predominantly on technology developed in the United States.

Rees, who was appointed to the advisory role by the UK finance ministry in January, warned that the country has only a narrow window in which to reinforce its competitive position in artificial intelligence. Speaking in an interview with Reuters, she stressed the urgency of taking decisive action, stating, “Now is the time for us to think very strategically about what we need to do to protect our leading position moving forward. Time really is of the essence; we don’t have two years here.”

Anthropic, the U.S.-based AI company, introduced Mythos in April to a limited group of organizations, with U.S. banking giant JPMorgan among the early recipients. The model has attracted significant attention within the banking industry because it is widely regarded as the most sophisticated AI system currently available for identifying cybersecurity vulnerabilities. Financial institutions view the technology as particularly valuable because it enables them to detect weaknesses in their systems more rapidly, implement security patches more efficiently, and strengthen their cyber defenses against increasingly sophisticated threats.

Despite the demand for the technology, access within the United Kingdom has remained highly restricted. According to Rees, only a very small number of banks operating in Britain have been granted access to Mythos, and those institutions are predominantly the UK subsidiaries of American banking groups. Major British lenders, by contrast, continue to wait for access without any clear indication of when it may become available. The chief executive of one of the UK's largest banks confirmed last week that the country's leading domestic banks still have no timeline for when they expect to gain access to the model.

The restrictions stem from measures introduced by the U.S. administration governing who may use Mythos. While those controls initially limited access significantly, Anthropic announced at the end of June that it had received authorization to release the AI model to trusted organizations within the United States. The company has since indicated that international expansion has begun. An Anthropic spokesperson said, “We have begun the rollout of Mythos 5 to organisations outside the United States. We continue to coordinate with the US government to expand access to the broader set of domestic and international partners.”

The issue has also attracted attention from the Bank of England. During his Mansion House speech on Tuesday evening, Governor Andrew Bailey renewed his appeal for greater international cooperation in managing the risks associated with frontier artificial intelligence technologies. He argued that effective oversight cannot be achieved by individual countries acting independently because advanced AI systems operate across borders. Bailey stated, “This needs to be coordinated internationally because no country can seal itself off from the cross-border nature of systems that are prevalent today.”

Alongside these concerns, Rees and fellow government AI adviser Rohit Dhawan, who serves as Head of AI at Lloyds Banking Group, have prepared a comprehensive package of policy and regulatory recommendations aimed at accelerating artificial intelligence adoption throughout the UK's financial services industry. Their proposals were released on Tuesday as part of the government's Financial Services AI Adoption Plan, published in advance of Chancellor Rachel Reeves' annual Mansion House address.

Among the recommendations is a proposal for regulators to assess the rapidly expanding use of AI-powered chatbots that provide financial guidance and advice to consumers. The advisers argue that increased regulatory attention will be necessary as these technologies become more deeply integrated into financial services. The UK government has stated that it intends to work closely with regulators and industry participants to determine the next steps for implementing the recommendations.

In a statement accompanying the publication of the AI Adoption Plan, Rachel Reeves reaffirmed the government's commitment to strengthening Britain's technological independence. She said, “set out a serious plan for AI sovereignty, backing ⁠British companies to compete and win”, adding that the Financial Services AI Adoption Plan forms a central part of delivering those ambitions across the financial sector.

Financial regulators have consistently warned that increasing reliance on a small number of major technology providers presents both operational and financial stability risks. Concentration among critical suppliers could leave financial institutions vulnerable if disruptions occur or if access to essential technologies becomes restricted.

Rees believes that, in addition to expanding domestic AI capabilities, Britain should broaden its strategic options by developing stronger relationships with artificial intelligence companies based outside the United States, including firms operating in countries such as China and France. Diversifying partnerships, she argued, would reduce dependence on any single jurisdiction while increasing resilience within the UK's financial technology ecosystem.

Dhawan also proposed measures to improve transparency and strengthen oversight of the AI providers that banks already depend upon. He suggested that the UK could extend financial regulatory supervision to major artificial intelligence companies by formally designating them as "critical" service providers to the financial sector. Such a move would place them within the regulatory framework already being expanded to include four major U.S. cloud service providers announced on Friday, helping ensure greater accountability and resilience among the technology companies supporting Britain's banking industry.

By fLEXI tEAM