AI safety and research company Anthropic has confirmed that it has launched a probe into a report that said that unauthorised users have accessed to its Mythos model.
The Mythos model, which had not been made accessible to the public yet, poses a threat to cybersecurity according to Anthropic.
“We're investigating a report claiming unauthorised access to Claude Mythos Preview through one of our third-party vendor environments,” said Anthropic.
The investigation comes after reports by Bloomberg News earlier in the day said that a group of people had gained access to the model.
Bloomberg said a “handful” of users in a private online forum gained access to Mythos on the same day Anthropic said it was being released to a small number of companies including Apple and Goldman Sachs for testing purposes.
The report highlighted that the anonymous users got to Mythos through internal access that one of them had as a worker at a third-party contractor for Anthropic and also by deploying methods used by cybersecurity researchers.
The group has not run cybersecurity prompts on the model and is more interested in fiddling with the technology than causing trouble, Bloomberg cited in its report. The report corroborated the claims via screenshots and a live demonstration of the model.
Even so, news of the probable breach will be alarming to authorities who have raised concerns about Mythos's potential to cause turbulence. In addition to this, they will most likely look for answers as to how potentially damaging technology can be kept out of the wrong hands.
Kanishka Narayan, the UK's AI minister, was cited by the Guardian as saying that UK businesses “should be worried” about the model's ability to spot flaws in IT systems – which hackers could then act upon.
The model has been vetted by the world's leading safety authority for the technology, the UK's AI Security Institute (AISI), which warned last week that Mythos advances from previous models in terms of the cyber-threat it posed.
AISI underlined that Mythos could execute attacks that required multiple actions and discover weaknesses in IT systems without human intervention. Moreover, it highlighted that these tasks would normally take human professionals days to carry out.
Mythos was the first AI model to successfully complete a 32-step simulation of a cyber-attack created by AISI, solving the challenge in three out of its 10 attempts.
Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories — On NDTV Profit.
