Anthropic’s Mythos AI Reportedly Hacked the NSA’s Most Sensitive Systems ‘in Hours’

The cybersecurity world was shaken when Anthropic's AI model, Mythos, was reported to have allegedly hacked the NSA's most sensitive systems in hours. This claim originated from a report by The Economist citing Senator Mark Warner, who said he received the information from NSA General Joshua Rudd. The report sparked panic and was seen as evidence that the worst fears about Mythos were true. However, a subsequent report from The New York Times clarified that the NSA's internal tests with Mythos were conducted in a robustly controlled digital environment, making it highly unlikely that any hacker or foreign intelligence agency could replicate them. The report also stated that while Mythos was able to identify cybersecurity vulnerabilities, it did not actually exploit them. Furthermore, the author of the initial report admitted that his portrayal of the tests had been misleading, noting that he quoted Senator Warner to give a sense of Mythos' potency but failed to add necessary caveats. This situation has also led to a conflict with the Trump administration, which ordered a ban on access to Mythos-class models for foreign nationals, citing national security concerns. The administration's move was criticized by many cybersecurity experts, who argued it would hamper U.S. cybersecurity defenses. The ban reportedly put a halt to the NSA's internal tests with Mythos, but the administration is now working with Anthropic to reinstate the agency's access for limited purposes related to national security.

(Source：Google News)