UK banks blocked from Mythos as 2 regulators flag AI risks

Britain's largest banks cannot access Anthropic's most advanced AI model, a gap a government adviser called a strategic vulnerability requiring urgent sovereign investment.

UK banks have been blocked from accessing Anthropic's Claude Mythos, the startup's most powerful artificial intelligence model, a restriction a government-appointed banking industry adviser called a wake-up call for Britain to build its own AI capabilities.

"The inability of UK banks to access Mythos is a strategic vulnerability that demands an urgent national response," the adviser said in a statement reviewed by Reuters. The adviser, whose role focuses on technology strategy in the banking sector, said Britain must adopt a more coordinated approach to AI development or risk falling behind in financial technology.

The warning comes as Canada's Office of the Superintendent of Financial Institutions separately cautioned domestic banks about risks tied to Claude Mythos and other advanced AI models, according to a July 13 report from Reuters. The Canadian regulator said the technology could increase cyber threats and reduce the time available for banks to detect and respond to attacks.

Anthropic's Claude Mythos represents a step-change in AI capability, with performance benchmarks that exceed prior models across reasoning, coding, and multilingual tasks. The model's restricted availability has created a two-tier system where US financial institutions maintain access while their UK counterparts do not, the adviser said.

The Sovereign AI Imperative

The adviser's call for domestic AI investment aligns with broader government efforts to position Britain as a leader in AI safety and development. The UK hosted the first global AI safety summit in November 2023 at Bletchley Park and has committed 100 million pounds to establish a Foundation Model Taskforce, later renamed the AI Safety Institute.

However, the adviser argued that funding and institutional focus remain insufficient relative to the pace of US and Chinese AI advancement. Without dedicated compute infrastructure and model development capacity tailored to financial services, UK banks will remain dependent on foreign AI providers whose access policies can shift without notice, the adviser said.

The UK banking sector processes more than 8 billion transactions annually and employs over 400,000 people in finance and technology roles, according to UK Finance data. The sector's reliance on AI for fraud detection, risk modeling, and customer service has grown sharply, with adoption accelerating after the release of large language models in 2023.

Cross-Border Regulatory Divergence

The divergence between Canadian and UK regulatory approaches highlights the fragmented global response to advanced AI in finance. Canada's OSFI issued its warning as a precautionary measure, urging banks to assess their exposure to models whose behavior may not be fully predictable. The UK has not issued a similar directive, instead focusing on enabling access through its pro-innovation regulatory framework.

Anthropic, which has raised more than $7 billion from investors including Google and Amazon, has not publicly commented on geographic access restrictions for Mythos. The company has stated that it evaluates deployment on a case-by-case basis, prioritizing safety and alignment with its responsible AI framework.

The adviser said the next 12 months will be critical for UK AI strategy, with decisions on compute infrastructure investment and model development expected before the end of 2026. Without action, the gap between UK and US banks' AI capabilities will widen, the adviser said.

This article is for informational purposes only and does not constitute investment advice.