Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell urgently met with top Wall Street leaders to discuss the cyber risks posed by Anthropic’s new, more powerful AI model, which could be exploited for sophisticated attacks on critical financial infrastructure. In response, Anthropic has restricted access to the AI through a collaborative initiative called Project Glasswing with trusted partners to proactively secure vital systems, highlighting a shift toward responsible AI development focused on balancing innovation with cybersecurity.
Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened an urgent meeting with top Wall Street leaders to discuss concerns about the latest AI model released by Anthropic. The meeting, held at the Treasury’s headquarters in Washington, was arranged on short notice and included chiefs from major banks such as Citigroup, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs. JPMorgan’s Jamie Dimon was unable to attend. These banks are considered systemically important, highlighting the critical nature of the discussion for global financial stability.
The primary focus of the meeting was to raise awareness among these financial leaders about the potential cyber risks posed by Anthropic’s new AI model. This model is more powerful than previous versions and has demonstrated significant improvements in coding and reasoning capabilities. Regulators are increasingly concerned that this advanced AI could be exploited to launch sophisticated cyberattacks, posing a serious threat to the financial industry and beyond.
Anthropic’s new AI system is capable of identifying and exploiting vulnerabilities across major operating systems and web browsers when directed by a user. This capability has heightened fears that malicious actors could leverage the technology to compromise critical infrastructure. In response, Anthropic has limited access to the model, sharing it only with a select group of trusted partners, including Amazon, Apple, and JPMorgan, under a collaborative initiative called Project Glasswing.
Project Glasswing aims to allow these trusted companies to work proactively to secure vital systems before similar AI models become widely available. This cautious approach by Anthropic reflects an awareness of the potential dangers associated with releasing such powerful technology without adequate safeguards. It also underscores a broader shift in the AI development landscape, where progress is increasingly measured not just by new capabilities but also by responsible management and containment of risks.
Overall, this meeting and Anthropic’s measured release strategy signal a new phase in the AI race, emphasizing the importance of cybersecurity and risk mitigation. Regulators and industry leaders are now grappling with how to balance innovation with safety, as AI technologies become more capable and potentially more dangerous. The financial sector, given its systemic importance, is at the forefront of these efforts to anticipate and prevent emerging cyber threats linked to advanced AI.
