Anthropic’s Mythos AI Sparks Global Calls for Broader Access and Oversight

Anthropic’s latest AI model, Claude Mythos, has ignited global debates over cybersecurity and AI governance, as its unprecedented ability to uncover thousands of previously unknown software vulnerabilities leaves governments and businesses outside the U.S. clamoring for access. The model identifies 'zero-day' bugs, critical weaknesses that hackers could exploit to infiltrate systems, prompting fears of widespread cyberattacks on global infrastructure.

Anthropic has limited Mythos’ release to a select group of U.S.-based tech firms and is cooperating with the U.S. government, granting federal agencies initial access. While this cautious rollout aims to mitigate risks, it has raised concerns about the concentration of power in the hands of a single U.S. company. Critics argue that private entities like Anthropic should not unilaterally control technology with global implications.

'It doesn’t make sense that private individuals are deciding the fate of infrastructure for everyone else,' said Yoshua Bengio, a Turing Award-winning computer scientist and AI pioneer. 'What about all the companies and all the countries that didn’t get access?'

Internationally, the situation has fueled demands for 'AI sovereignty,' with many nations seeking independent AI capabilities to reduce reliance on U.S. tech. At the IMF and World Bank spring meetings in Washington, policymakers warned that Mythos could expose vulnerabilities in the global banking system, while European regulators expressed frustration over their lack of access to the model. The Bank of England, for example, secured assurances from Anthropic that U.K. banks would gain access this week.

The U.S. government is also taking steps to secure its own access, with the White House Office of Management and Budget instructing federal agencies—including Defense, Treasury, and Homeland Security—to begin using a version of Mythos. As tensions escalate, the Mythos rollout underscores the urgent need for international cooperation on AI governance to ensure equitable access and safeguard global security.