“Many of the new partners are vendors – companies or non-profits that maintain codebases that are relied upon by lots of other organisations around the world, including governments,” the company said in a statement released today (3 June).

“What each partner has in common is that a successful attack on their codebase could be catastrophic.”

The Treasury, the Systems of National Significance, which oversee critical infrastructure in the country, and the Reserve Bank of Australia, were already briefed on Mythos by Anthropic before access was offered.

Now, Australia’s national security firms are likely to be some of the first to get access to Mythos, according to reports by The Financial Times, which said that Anthropic announced that Five Eyes intelligence-sharing nations like Australia, as well as NATO and ENISA nations, would have access.

Meanwhile, OpenAI is offering its Mythos alternative, GPT-5.5 Cyber, to nations that are still awaiting their turn to try Anthropic’s model.

Most recently, the company has offered UK and Japanese banks access to the next iteration, GPT-5.5 Cyber, for use in cyber defence, an offer that Japan has taken the company up on.

Japan’s Finance Minister Satsuki Katayama said the offer was accepted after talks in Tokyo with OpenAI chief strategy officer Jason Kwon, adding that early access to the model will allow them to gain a hold against cyber criminals using advanced strategies and AI technology.

While Katayama did not name the financial institutions involved, she said that the move was “a big step forward in strengthening Japanese financial institutions’ ability to defend against cyber attacks”.

That being said, Katayama also said that both the Japanese government and the nation’s financial institutions expected to gain access to Mythos.

OpenAI also offered GPT-5.5 access to nine major UK banks, as Anthropic denied them access to Mythos. Banks include Lloyds Banking Group, HSBC, and Nationwide. Existing agreements mean that Santander and NatWest already have access.