UK regulators assess risks of new Anthropic AI model
Anthropic has triggered urgent risk assessments in the United Kingdom after regulators raised concerns about potential cybersecurity vulnerabilities linked to its latest artificial intelligence model.
British authorities are now working to evaluate possible threats posed by the system across critical financial infrastructure, News.Az reports, citing Reuters.
According to reports, the Bank of England, the Financial Conduct Authority, and the UK Treasury are holding discussions with the National Cyber Security Centre.
RECOMMENDED STORIES
The talks also involve major banks and financial institutions, as officials assess how the model could potentially expose vulnerabilities in sensitive IT systems.
A briefing for banks, insurers, and financial exchanges is expected in the coming weeks.
The scrutiny centers on Anthropic’s newest AI system, reportedly referred to as “Claude Mythos Preview,” which is part of a controlled testing initiative involving select organizations.
Officials are particularly concerned about whether advanced AI capabilities could be exploited to:
- Identify system weaknesses
- Assist cyberattacks
- Impact critical financial infrastructure
Anthropic has stated that the model is being tested under a program known as “Project Glasswing,” designed to explore cybersecurity applications in a controlled environment.
The company claims the model has already been used to detect thousands of vulnerabilities in widely used software, including operating systems and browsers.
The concerns in the UK follow similar discussions in the United States, where financial authorities have also begun examining potential risks from advanced AI systems.
Regulators say they are acting early to ensure that innovation in artificial intelligence does not create new systemic risks for global financial systems.
By Aysel Mammadzada
Similar news
