Anthropic’s Mythos AI Model Reportedly Accessed by Unauthorized Users as Bank Rollout Expands
Anthropic’s new Mythos AI model has reportedly been accessed by a small group of unauthorized users, according to Bloomberg News, which cited documentation and a person familiar with the matter. The company has said Mythos is powerful enough to help enable dangerous cyberattacks.
The report comes as Anthropic is also preparing to expand access to Mythos for European banks soon, after some large U.S. banks had already received initial access. The model has drawn attention from the cybersecurity and banking sectors because of its focus on cyber-related use cases.
Mythos is Anthropic’s new AI model focused on cybersecurity-related tasks.
Anthropic says the model is powerful enough to help enable dangerous cyberattacks, so unauthorized access raises security concerns.
Some large U.S. banks have already had initial access, and Anthropic plans to give access to European banks soon.
The source text does not include a direct statement from Anthropic on the report.
Reuters
TechCrunch
Bloomberg Markets and Finance
CNBC Television
Business Insider
The Guardian
Invezz
PYMNTS
WSJ
NYTimes
Bloomberg Technology
New York Post