Canada has access to Mythos AI’s powerful Anthropic model, says minister – National
Canada’s intelligence minister said Tuesday the federal government has access to a model of Anthropic Mythos — an AI platform so powerful the company has withheld it from the general public.
AI Minister Evan Solomon said the government had signed up to Project Glasswing, launched by Anthropic earlier this year to allow the US government and a limited number of high-tech companies to use Mythos to scan their systems for security risks.
Anthropic said earlier on Tuesday it was expanding the program to 150 other organizations in more than 15 countries.
“I can confirm that the Canadian government is now part of Project Glasswing, which allows companies to acquire Mythos,” Solomoni told reporters after a cabinet meeting in Ottawa.
“So we have access, we’re part of Project Glasswing, I’ll confirm that, and with the Canadian Cyber Security Centre, we have access to Mythos now. So it’s a very important step for Canadians and our government to make sure that we have access and we can protect our institutions and our countries.”
Solomon said he could not provide information on which companies can acquire Mythos.
Anthropic also declined to name specific companies it has access to but said it targets industries such as health, energy, water, telecommunications and hardware. This expansion also includes government agencies, the company said.
When announcing the creation of Mythos this spring, Anthropic said the model was “so good” at hacking and cybersecurity that it could release it to a small group of trusted organizations.
The company said that Project Glasswing was created with the hope of protecting the world’s sensitive software from a “bad” new model that could cause public safety, national security and the economy – especially if used in a major cyberattack.
Get the latest country news
Get the best Canadian news delivered to your inbox as it happens so you never miss a trending story.
Solomon said at the time that withholding Mythos from the public “was a responsible approach and gives people who protect sensitive programs a good head start.”
Anthropic said on Tuesday that its partners using Mythos so far – including Apple, Google, Amazon and Microsoft – have found more than 10,000 security flaws they consider serious or critical, although some industry experts say fears of unrestricted hacking are overblown.
Project Glasswing aims to solidify a large software infrastructure before powerful AI like Mythos is widely available from other providers, which Anthropic says is between six and 12 months away.
The company said it intends to bring Mythos-class models to all its customers, with added protection, in the coming weeks.
The expansion of the Mythos release comes after Anthropic privately filed what would be the largest initial public offering in history.
Solomon says AI strategy to emphasize privacy, build ‘trust’
The announcement of Canada’s involvement in testing Mythos comes ahead of the expected release this week of the federal government’s new AI strategy, which Solomon said Tuesday will emphasize increasing Canadians’ trust in technology.
“There’s going to be a lot of confidence, a lot of empowerment, building Canada. You’ll see the details later this week,” he told reporters.
Ottawa has previously said the long-delayed strategy will include new privacy and internet security rules the government will introduce to address concerns about AI.
Solomon said he is working with Culture Minister Marc Miller and Justice Minister Sean Fraser on legislation he is leading to revise Canada’s private sector privacy laws. Miller is spearheading a new cybercrime bill that would cover AI conversations, which Solomon talked about.
Solomon would not say whether those laws will be introduced before Parliament adjourns for the summer in less than three weeks.
“Over time, you’ll see, but I would just say that it’s a very important thing to build trust in the AI strategy and other things,” he said. “We have to protect Canadians first and foremost.
“We’re going to have a law, a comprehensive law, that will make sure we protect … first and foremost the children, the privacy of Canadians and the data of Canadians.”
Other expected features of the strategy include independent computing infrastructure, support for Canadian AI companies and international networking. The strategy is expected to provide AI training and education to Canadians.
Solomon also said it will consider the impact of technology on the labor market.
-via files from The Canadian Press, Associated Press and Reuters
© 2026 Global News, a division of Corus Entertainment Inc.
