Ireland's National Cyber Security Centre (NCSC) said it has reviewed the published technical material relating to a powerful new artificial intelligence model from AI company Anthropic.
Claude Mythos Preview was launched last week, but the company stopped short of a broad release, citing concerns it could expose previously unknown cybersecurity vulnerabilities.
Anthropic said the model is capable of identifying and exploiting weaknesses across "every major operating system and every major web browser".
The tool is currently only available to a restricted consortium of around 40 technology companies amid fears that it could be used as a powerful cyber weapon in the wrong hands.
In its update, the NCSC said the capabilities described by Anthropic appear to represent a significant change in the way hardware and software vulnerabilities are identified and patched.
"At present the advantage is with cyber defenders," the NCSC said.
"These new capabilities are in the hands of major technology vendors who are using them to identify and patch vulnerabilities in the hardware and software that underpins critical infrastructure worldwide."
"There is no indication that a comparable autonomous vulnerability discovery capability is available to threat actors at this time," the centre said in a statement.
The NCSC added that organisations should expect accelerated vulnerability disclosure over coming months as these tools are applied at scale.
"Anthropic's decision to restrict the release of the model and to work collaboratively with industry partners is a responsible approach," the NCSC said.
