Five Eyes Warns of AI-Armed Cyberattacks Within Months

[Financial News]  

The security chiefs of the Five Eyes intelligence alliance warned that cyberattacks armed with artificial intelligence from adversaries such as China and Russia could hit within months.
They said the West still holds the advantage for now, but warned that adversaries will develop attack methods with AI help that can break through those defenses. Five Eyes said the timing would be measured in "months, not years."
The Financial Times reported on the 22nd, local time, that Five Eyes issued the warning in a joint statement.
Five Eyes is a five-nation intelligence-sharing network led by the United States and involving the United Kingdom, Canada, New Zealand and Australia.
The warning came a week after the U.S., citing national security concerns, told AI startup Anthropic to block foreign access to its most advanced AI models, MisoS 5 and Fable 5.
The statement acknowledged that improvements in the capabilities of so-called frontier models are helping Western countries protect their core infrastructure. Still, it also expressed concern that AI is helping develop offensive cyber capabilities.
The cybersecurity chiefs said, "Frontier AI models are expected to fundamentally transform both cyberattack and cyberdefense capabilities, surpassing current industry expectations."
The statement did not name adversaries such as China or Russia, but stressed that Western governments and companies must prepare for more powerful and sophisticated cyberattacks designed with existing AI models.
Google Threat Intelligence Group said last month that an unidentified attacker had tried to use AI to exploit a previously unknown vulnerability, but the attempt was stopped in advance. Google also said Russian hackers were believed to be abusing AI models to launch attacks that automatically adapt to security defenses. According to Google, hacking groups in China and North Korea are making similar attempts.
The capabilities of Western frontier AI models such as Claude MisoS are well known, but the models built by China and Russia remain shrouded in secrecy.
In the statement, the Five Eyes cybersecurity chiefs urged Western companies to adopt AI for cyber defense. "Those who delay adopting AI ... will face increasing risks," the statement warned.

dympna@fnnews.com Song Kyung-jae Reporter