In brief
- U.S. officials warned major banks about cybersecurity risks tied to Anthropic’s Mythos AI model, Bloomberg reports.
- The system can reportedly identify and exploit vulnerabilities in operating systems and browsers.
- Anthropic has limited access to the model while it evaluates potential security risks.
U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell reportedly convened a meeting with Wall Street bank CEOs earlier this week to warn about cybersecurity risks tied to a new artificial intelligence model from Anthropic.
According to a report by Bloomberg, the meeting included executives from Citigroup, Bank of America, Wells Fargo, Morgan Stanley, and Goldman Sachs. Officials discussed Anthropic’s new AI model Mythos, which has recently drawn broad concern over its apparent advanced cybersecurity capabilities.
Officials convened the meeting to ensure banks understand the risks posed by systems capable of identifying and exploiting software vulnerabilities across operating systems and web browsers, and to encourage institutions to strengthen defenses against potential AI-assisted cyberattacks targeting financial infrastructure.
Security researchers have warned that tools capable of automatically discovering vulnerabilities could accelerate both defensive security work and malicious hacking if misused.
Anthropic’s Mythos model first surfaced online in March after draft materials about the system leaked online, revealing what the company described as its most capable AI model yet. In testing, the system reportedly found thousands of previously unknown software vulnerabilities, including zero-day flaws across major operating systems and web browsers.
Anthropic researchers said in a report earlier this week that Mythos Preview’s vulnerability-discovery capabilities were not intentionally trained, but instead emerged from broader improvements in the model’s coding, reasoning, and autonomy.
“The same improvements that make the model substantially more effective at patching vulnerabilities also make it substantially more effective at exploiting them,” the firm wrote.
Because of those capabilities, Anthropic has restricted access to a small group of cybersecurity organizations.
“Given the strength of its capabilities, we’re being deliberate about how we release it,” Anthropic said in a statement. “As is standard practice across the industry, we’re working with a small group of early access customers to test the model. We consider this model a step change and the most capable we’ve built to date.”
To address that risk, Anthropic is testing Mythos through Project Glasswing, a collaboration with major technology and cybersecurity companies that uses the model to identify and patch vulnerabilities in critical software before attackers can exploit them.
“Project Glasswing is a starting point. No one organization can solve these cybersecurity problems alone,” the company said in a statement. “Frontier AI developers, other software companies, security researchers, open-source maintainers, and governments across the world all have essential roles to play.”
Anthropic did not immediately respond to Decrypt’s request for comment.
Disclaimer: The information on this page may come from third parties and does not represent the views or opinions of Gate. The content displayed on this page is for reference only and does not constitute any financial, investment, or legal advice. Gate does not guarantee the accuracy or completeness of the information and shall not be liable for any losses arising from the use of this information. Virtual asset investments carry high risks and are subject to significant price volatility. You may lose all of your invested principal. Please fully understand the relevant risks and make prudent decisions based on your own financial situation and risk tolerance. For details, please refer to
Disclaimer.
Related Articles
China to Strengthen Legislative Framework for AI, Digital Economy, and Blockchain
China plans to enhance its network legal system to address the needs of internet development, focusing on legislation for AI, digital economy, and other tech sectors. By December 2025, over 180 laws are expected, though current regulations still lag behind public expectations.
GateNews27m ago
Intchains Group Discloses 8,040 ETH Staked, Accelerates AI-Powered Operations Transformation
Intchains Group reported a cumulative staking of 8,040 ETH, with a strategy that includes the FalconX platform and its own Goldshell Stake. The company is also enhancing operations through AI automation to boost efficiency and decision-making.
GateNews57m ago
Adata and Jay Chou raking in big money? GD’s talent agency secretly meets with Dubai’s royal family to discuss an IPO, AI, and robots
South Korean entertainment technology company Galaxy Corporation has recently moved into the Middle East market, established the Dubai entity “Galaxy ME,” and met with members of the UAE royal family to explore the integration of AI and content. The company has selected an underwriter to carry out its IPO; last year, revenue reached 300 billion Korean won, and it has attracted investments from Jay Chou and Adata, among others. This move is intended to drive the development of the Middle East entertainment ecosystem and underscores the importance of its globalization strategy.
ChainNewsAbmedia2h ago
China team EvoMap erupts in anger over Hermes Agent copying amid viral success: the self-evolving system has a high degree of similarity
The China team, EvoMap, accuses Nous Research’s Hermes Agent of copying its Evolver/GEP architecture, claiming that the two are highly similar across multiple key modules. EvoMap says that Hermes’ design does not reference or acknowledge its previously released content. The controversy has sparked widespread attention in the community. Some Chinese netizens have voiced support for EvoMap and criticized Hermes’ copying behavior. Although the Nous Research team has not yet rebutted the allegations point by point, the incident has quickly spread across AI public discourse.
ChainNewsAbmedia2h ago
OpenAI early investors are bullish! Brain-machine startup Sabi plans to ship by the end of this year with a “thinking-to-type” beanie
Silicon Valley startup Sabi announced that it is developing a non-invasive wearable brain-computer interface that will let users input text into a screen just by “thinking.” It is expected to launch by the end of 2026. The device will be equipped with up to 100,000 micro-sensors, helping to accurately interpret brain activity, and Sabi also emphasizes data privacy protection.
ChainNewsAbmedia2h ago
Data reveals that “Claude got less intelligent” isn’t an urban legend; an AI model’s instability can become a corporate risk.
The article discusses the phenomenon of unstable performance of LLMs (large language models) in real-world applications by AI companies, calling it “de-intelligence,” and provides examples to illustrate its real impact on business workflows. Data shows that most mainstream models are currently in a degraded state, affecting companies’ productivity and stability. Companies need to start treating model stability as a new standard; otherwise, they will face infrastructure risks.
ChainNewsAbmedia3h ago
