Pursuing its powerful Mythos model, Anthropic announced its launch earlier this week but refrained from a full release, citing concerns over potential undiscovered cybersecurity vulnerabilities.
The company stated that the model is designed to identify and exploit weaknesses across “every major operating system and every major web browser”.
Last week, Anthropic mentioned that it was engaged in discussions with US government officials regarding the model’s “offensive and defensive cyber capabilities.”
A third source close to the issue confirmed Anthropic’s proactive outreach, indicating the company briefed senior US government officials and key industry players about Mythos’s functionalities before its release.
The Treasury-hosted meeting in Washington on Tuesday aimed to ensure banks recognized the risks posed by Mythos and similar models while taking necessary steps to secure their systems, as noted by one source.
Also Read: Royal Enfield just launched a new motorcycle, but Nomura wants some refinement
Invitations were sent out while the majority of CEOs from the largest US banks were already in Washington for other engagements, according to one source.
Access to Mythos will be restricted to around 40 technology companies, including Microsoft and Google, as stated by the startup.
Bloomberg News, the first outlet to report on the matter on Thursday, indicated that the CEOs of Citigroup, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs were in attendance. JPMorgan CEO Jamie Dimon was unable to participate, one of the sources informed Reuters.
Goldman Sachs and the Federal Reserve opted not to comment. The Treasury, lenders, and Anthropic did not immediately respond to Reuters’ requests for commentary.
