Five Eyes intelligence warns of AI for devastating cyberattacks emerging within months

Five Eyes intelligence warns of AI for devastating cyberattacks emerging within months

June 23 2026, 05:47 AM • 1376 views

The intelligence services of the Five Eyes countries warned in a joint statement that advanced AI models will emerge in months, not years, accelerating cyber threats. They called on leaders to act now, emphasizing that cyber risk has become a primary business risk.

Powerful artificial intelligence models capable of devastating new cyberattacks on governments and businesses will emerge in just a few months, intelligence agencies of the Five Eyes countries warn in a rare joint statement, urging leaders to "act now," writes UNN citing The Guardian.

The unexpected public intervention by the signal intelligence agencies of Australia, the United States, the United Kingdom, New Zealand, and Canada comes after the Trump administration earlier this month decided to ban "foreign nationals" from using the highly publicized artificial intelligence model created by the technology company Anthropic, called Fable.

The statement says that while artificial intelligence "will help us improve cyber defense over time, it also accelerates the speed, scale, and complexity of cyber threats."

Advanced artificial intelligence models are expected to surpass current industry expectations, fundamentally transforming both offensive and defensive cyber capabilities. The timeline is not years, but months

"In this environment, cyber resilience is an integral part of promoting business continuity, market confidence, and long-term value," the statement said.

The cybersecurity agencies stated that leaps in artificial intelligence models have shown that this technology will lower barriers for attackers and increase the speed and complexity of attacks.

"A whole-of-organization and whole-of-society response is needed," the statement continues. Five Eyes is an intelligence alliance formed by five countries after World War II.

"Cyber risk can no longer be viewed as a purely technical issue. It is a core business risk and a management responsibility," the statement notes.

Generative artificial intelligence models are powerful new tools capable of searching for vulnerabilities in cybersecurity systems, and they can help exploit those vulnerabilities as well as fix them.

"What is different about the latest [AI models] is that they are very good at generating exploits," said Olivia Shen, an expert on national security and artificial intelligence at the United States Studies Centre at the University of Sydney.

Although the Five Eyes statement does not mention any specific AI models or companies, many around the world have taken note of the cutting-edge level of Anthropic's tools.

One of the latest inventions of the major technology company is called Fable 5, supposedly a more community-friendly version of Mythos—a powerful artificial intelligence model released earlier this year, capable of detecting vulnerabilities in cyber systems, which is only available to verified organizations and companies due to concerns about its potential misuse.

Both Anthropic models were suspended for use by "foreign nationals" in June by the U.S. government, which cited recommendations from national security authorities.

Shen said much of the world is focused on what will happen next with Anthropic, but many more powerful AI models may not be far off.

"I think we should anticipate that the next Mythos or the next Fable is not far away," Shen said.

"We can only see what has been released, but there may be other models being developed by equally advanced countries like China, or by other states and other entities and companies," she added.

Recently, a claim spread online that Anthropic's Mythos model had breached almost all secret systems of the U.S. National Security Agency (NSA).

The claim came from Senator Mark Warner, vice chairman of the Senate Intelligence Committee. The Economist reported his account of what the NSA director told him.

Warner stated that General Joshua Rudd, who heads the NSA and U.S. Cyber Command, described the tool in stark terms.

"This tool broke almost all of our secret systems not in weeks, but in hours," The Economist wrote, citing Warner.

Warner cited this example to praise Anthropic, not to condemn it. He used it to justify the need for faster pre-testing of new models.

However, as noted in specialized media, the detail missing from the network is simple. It was an authorized "red team" test on the agency's networks, not an external intrusion.

Shashank Joshi, the Economist editor who published the quote, later cautioned against taking it literally. He said Mythos's effectiveness depends on it working in conjunction with other tools under specific conditions.

The U.S. government was already a partner of Mythos. Anthropic deployed this model for government cyber defenders as part of Project Glasswing back in April.

Analyst Kyle Chase noted that the breach was a test. He said the real trigger was a separate jailbreak flagged by Amazon.

Anthropic's statement confirms this. It says the mentioned jailbreak simply asked the model to read code and fix errors.

This method uncovered several minor, already known errors that competing models, such as OpenAI's GPT-5.5, could also detect.