NZ gets access to hacking Mythos AI as Trump shores up national security on AI

New Zealand's cyber watchdog has been given access to a frontier AI model with worrisome hacking powers at the same time as Donald Trump moves to shore up US national security against any threats.

Anthropic, the creator of the Mythos AI, has begun distributing it to 150 organisations worldwide beyond the smaller US group it gave access to in April.

The National Cyber Security Centre here said it had received access on behalf of the New Zealand government.

"The NCSC already has productive working relationships with a range of international partners and vendors, including some who have had access to preview models, but direct access will strengthen our national cyber security mission," it told RNZ on Thursday.

The US group with access has been testing Mythos's ability to break into systems.

Anthropic then let them share cybersecurity threats with others who might face similar vulnerabilities.

In the US, a second congressional Homeland Security hearing was scheduled for Thursday on frontier AI threats.

President Trump signed an executive order on Tuesday for vetting of advanced AI for national security risks.

"Advanced AI capabilities make our nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies," it said.

Trump pulled out of signing an AI order in May saying it might hold the US back.

The new order sought to "harden" information systems against the threats, and requested that "AI companies voluntarily provide the federal government access to 'covered frontier models' for a cybersecurity review up to 30 days before their planned release to 'other trusted partners'," reported the Council for Foreign Relations in the US.

The National Cyber Security Centre here said it welcomed Anthropic's provision of access to Mythos.

This would help with its response to cyber security risks and opportunities around frontier AI, including with advice to local organisations to help improve their cyber security.

"As frontier AI models improve, they will change the cyber threat landscape for organisations because of the ability for malicious actors to find and exploit vulnerabilities at unprecedented speed and scale. However, AI can also assist network defenders in their work and increase their ability to protect systems at scale and pace," the centre said.

Its advice was to patch software frequently, limit where an attack could get in, review how vulnerabilities are managed and monitor a lot.

Sign up for Ngā Pitopito Kōrero, a daily newsletter curated by our editors and delivered straight to your inbox every weekday.