Intelligence alliance "Five Eyes" warned about the potential of AI in cyber attacks on countries

Artificial intelligence models capable of hacking the defenses and information security systems of world governments will appear in a few months. This was announced on June 23 in a joint statement within the framework of the Five Eyes alliance by the intelligence services of the United States, Great Britain, Canada, Australia and New Zealand.

"Advanced artificial intelligence models are expected to exceed current industry expectations. <...> The timeline for this process is not years, but months," said a statement posted on the Australian government's website.

The security services have proposed a list of measures to counter hacking from artificial intelligence, as well as to develop tools to counter AI models. The alliance proposed to implement multi-level protection in all security systems and apply AI for defensive purposes.

More valuable than uranium: what will lead to the closure of the US access to the best AI model

The consequences of the decision will hit Europe

The Economist magazine reported on June 21 that the Mythos artificial intelligence model of the company Anthropic has caused concern in the United States after reports that it was able to hack secret systems of the National Security Agency (NSA). It is noted that possible restrictions on the use of advanced AI models abroad are being discussed in the United States. At the same time, such measures are already causing controversy about the difficulty of controlling the spread of powerful technologies.

On June 13, Anthropic announced that it had suspended access to its AI models Fable 5 and Mythos 5 after receiving an order from the US authorities, citing national security concerns. The company stressed that it complies with the requirements, but does not agree with the decision and expects to restore access to the services in the near future.