Unauthorized group gained access to Anthropic’s proprietary Mythos cyber tool, report claims

A group of unauthorized users has reportedly gained access to Mythos, Anthropic’s recently announced cybersecurity tool.

Much has been made of Mythos and its supposed power – an artificial intelligence product designed for enterprise security that, in the wrong hands, could become a powerful hacking tool, according to the company. Now, Bloomberg has reported that a “private online forum,” whose members have not been publicly identified, managed to gain access to the tool through a third-party vendor.

“We are investigating a report alleging unauthorized access to the Claude Mythos Preview through one of our third-party vendor environments,” an Anthropic spokesperson told TechCrunch. The company said that, so far, it has found no evidence that the alleged unauthorized activity has affected Anthropic’s systems in any way.

The unauthorized team tried a number of different strategies to gain access to the model, including using “access” enjoyed by the person interviewed by Bloomberg. That person is currently employed by a third-party contractor working for Anthropic, the agency said.

Members of the group are part of a Discord channel seeking information about unreleased AI models, the agency said. The team has been using Mythos regularly since gaining access to it and provided Bloomberg with screenshots and a live demo of the software.

Bloomberg reports that the team, which supposedly gained access to the tool the same day it was publicly announced, “made an educated guess about the model’s location online based on knowledge about the format Anthropic used for other models.” The group in question is “interested in playing with new models, not wreaking havoc with them,” the source told the agency.

Mythos has been released to a select number of vendors, including big names like Apple, as part of an initiative called Project Glasswing. The model’s limited release was designed to prevent bad actors from using it. The tool could weaponize against corporate security rather than strengthen it, Anthropic said.

If true, unauthorized use of Mythos could cause problems for Anthropic, which provided the proprietary version to assuage the company’s concern about enterprise security.