The Five Eyes international intelligence alliance has warned that artificial intelligence models capable of carrying out large-scale cyberattacks and overcoming government and corporate defenses are just months away. Intelligence agencies in the US, UK, Canada, Australia and New Zealand have urged businesses and governments to immediately step up their defenses against sophisticated digital threats.
The Five Eyes international intelligence alliance has warned that artificial intelligence models capable of carrying out large-scale cyberattacks and overcoming government and corporate defenses are just months away. Intelligence agencies in the US, UK, Canada, Australia and New Zealand have urged businesses and governments to immediately step up their defenses against sophisticated digital threats.
This was reported by the American television channel CNN, citing a joint statement by the intelligence alliance.
According to intelligence estimates, new models of artificial intelligence of the Frontier AI class will soon radically change both offensive and defensive potential in cyberspace. The main danger is that the technologies significantly lower the threshold of entry for attackers, as well as increase the speed and complexity of conducting digital attacks.
«Changes in artificial intelligence are rapidly transforming cyber risks, and we must act quickly to stay ahead,» the group of intelligence agencies said.
Olivia Shen, Director of the Strategic Technology Program at the Center for US Studies at the University of Sydney, emphasized in a media comment that in the era of AI, breaches and data leaks are becoming inevitable.
«The question is not whether they will happen, but when. Therefore, it is important to prepare now,» the expert noted.
She said that small and medium-sized businesses, which typically have limited cybersecurity budgets, are at the greatest risk. Unlike large corporations that systematically invest in defense, small companies risk becoming «easy prey» for hackers armed with AI.
To counter new threats, experts recommend that companies:
promptly update outdated systems and fix software vulnerabilities;
strictly limit the number of people who have access to critical infrastructure;
implement their own AI tools to monitor anomalous network behavior and respond quickly to incidents.
The intelligence warning comes amid the recent decision by the US administration to restrict foreign nationals from accessing advanced AI models Mythos 5 and Fable 5 from startup Anthropic. The Mythos model in particular has raised serious concerns due to its extraordinary ability to find vulnerabilities in the code. In addition, the developers discovered methods to bypass internal security barriers («jailbreaking») in the Fable model, which forced the company and the government to begin emergency consultations.
