Canada’s AI minister says Anthropic withholding Mythos is ‘responsible’ – National

Anthropic released a system preview card surrounding its newest AI model, Claude Mythos, which the company stated was “substantially beyond those of any model we have previously trained” and therefore would not be released to the public, citing cybersecurity dangers.

Get daily National news

Get daily Canada news delivered to your inbox so you’ll never miss the day’s top stories.

“Anthropic and the Canadian government are engaged in constructive, ongoing discussions,” Solomon said in an emailed statement from his office to Global News.

“I met with Anthropic this morning as part of our continued engagement with leading AI companies on safety, security, and Canada’s sovereign interests. The Government of Canada takes the protection of its systems, its critical infrastructure, and Canadians’ data with the utmost seriousness.

Solomon stated that “Anthropic’s approach of working with defenders first, rather than releasing this new model broadly, is the responsible path and gives people protecting critical systems a head start.”

Concerns about rapid AI development and ransomware have grown significantly amongst Canadians.

A January 2026 federal report by the Canadian Centre for Cyber Security stated that numerous Canadian organizations and businesses, “regardless of size or sector,” as well as individuals, are susceptible to ransomware attacks.

However, “critical infrastructure and large corporations” were found to be the top targets for ransomware activities.

&copy 2026 Global News, a division of Corus Entertainment Inc.