Solana news: Anthropic Expands Project Glasswing to UK Financial Institutions
Anthropic Extends Claude Mythos Preview Access
Anthropic has announced it will expand access to its Claude Mythos Preview, a frontier AI model, to UK financial institutions as part of Project Glasswing. This initiative is a controlled early-access programme focusing on defensive cybersecurity applications.
Project Glasswing Partners and Funding
Project Glasswing initially launched with 12 partner organisations, including major technology and security firms such as Amazon, Apple, Cisco, CrowdStrike, the Linux Foundation, Microsoft, and Palo Alto Networks. In total, 40 organisations have received access to the Mythos Preview. Anthropic has committed up to USD 100 million in usage credits and USD 4 million in direct donations to open-source security organisations to support these efforts.
Cybersecurity Capabilities and Financial Sector Impact
Recent testing of Claude Mythos Preview identified thousands of zero-day vulnerabilities in major operating systems, web browsers, and other critical software. Anthropic notes that these capabilities are a result of general improvements in AI reasoning and software engineering, not explicit training for software exploitation.
The UK's AI Security Institute has confirmed access to the model, reporting improved performance in simulating multi-step cyber attacks compared to previous AI models. The controlled release to UK financial institutions addresses the sector's exposure, given the mix of legacy and modern infrastructure in banking.
Regulatory and Industry Response
Bank of England Governor Andrew Bailey has called for rapid evaluation of the risks posed by advanced AI models. In the US, Treasury and Federal Reserve officials have convened discussions with financial leaders regarding the implications of Mythos and similar technologies.
By restricting access, Project Glasswing aims to give defenders time to identify and patch vulnerabilities before broader release of comparable AI models.
