Anthropic announced on Tuesday that it is significantly broadening access to its tightly controlled Mythos model, granting permissions to approximately 150 new organizations.
This expansion brings the total number of entities with access to the model to around 200.
Expanding Industry Representation
The AI lab stated that the newly added organizations, which were not named in the announcement, span various sectors that were "underrepresented in the initial user group."
These sectors include power, water, telecommunications, healthcare, and hardware industries.
Anthropic noted that many of the new partners are from over 15 countries and are suppliers of codebases relied upon by governments and other major national institutions.
Purpose of Access and Security Focus
Anthropic explained that granting these critical institutions access to the Mythos model is intended to allow them to leverage it to bolster their own cybersecurity defenses.
The model's release has raised concerns about the potential hacking capabilities of frontier AI models, with reports indicating Mythos possesses advanced cybersecurity and hacking abilities that could pose a serious threat to national security and other critical infrastructure.
The company stated, "We expect that in the next 6 to 12 months, many other AI companies will have models at the Mythos level, and they may release them without safeguards to prevent abuse. We hope the 'Project Glasswing' initiative will encourage organizations to establish operational norms that reflect this reality."
Initial and New Partners
The initial group of approximately 50 partners included the U.S. government, leading investment banks like JPMorgan Chase (NYSE: JPM) and Goldman Sachs (NYSE: GS), and a range of major tech companies such as Amazon.com (NASDAQ: AMZN), Apple (NASDAQ: AAPL), NVIDIA (NASDAQ: NVDA), Broadcom (NASDAQ: AVGO), and CrowdStrike Holdings, Inc. (NASDAQ: CRWD).
Anthropic reported that this first wave of partners has already discovered over 10,000 "high-severity or critical" security vulnerabilities.
Each new partner organization must pass a security review before gaining access to the model.
Anthropic wrote, "What all partners have in common is that the consequences of a successful attack on their codebases could be catastrophic. We estimate that for most partners, a major attack could affect over 100 million people and have significant global and national security implications."
Corporate Developments
Anthropic's CEO, Dario Amodei, recently appeared at an event hosted by the AI company.
Separately, the company confidentially filed paperwork for a public listing with the U.S. Securities and Exchange Commission on Monday, June 1st, paving the way for what is expected to be one of the largest and most high-profile initial public offerings this year, alongside SpaceX and fellow AI lab OpenAI.
Last week, the AI lab announced it had raised $650 billion in a Series H funding round, bringing its post-money valuation to $9.65 trillion.
Anthropic also stated that its recent annual recurring revenue has surpassed $470 billion.
Comments