Anthropic’s Most Advanced AI Models Shut Down Globally: US Government Orders Access Block for Foreign Nationals Due to Cyberattack Fears
AI startup Anthropic has shut down its most advanced AI models, ‘Claude Fable 5’ and ‘Mythos 5’, worldwide. The company took this decision following a US government order mandating a ban on the use of these models by foreign nationals.
Now, no non-US citizen—including Indians—will be able to use them. US officials cited national security concerns and the threat of cyberattacks as the reasons behind this move. This marks the first time a government has banned the use of AI software itself, rather than hardware or chips.
Fear of Cyberattacks via ‘Jailbreaking’
The US government fears that these advanced AI models could be ‘jailbroken’—meaning their security safeguards could be bypassed. If this were to happen, the models could be exploited to identify vulnerabilities in computer systems, hack government data, or launch major cyberattacks on banking systems and critical infrastructure.
According to a report by *The Wall Street Journal*, Amazon researchers managed to extract information about software vulnerabilities from Anthropic’s model during testing by using specific prompts. This information was reported to the US Department of Commerce, prompting immediate, strict action.
Company Calls the Ban a Mistake
Anthropic maintains that the government’s decision stems from a major misunderstanding. The company argues that the identified vulnerabilities were very limited and that other public AI models available on the market could uncover similar flaws; therefore, a sweeping global ban is unjustified.
The company stated that prior to launch, it had conducted weeks of testing with US government agencies and the UK’s AI Safety Institute, during which no such major risks were detected. Anthropic is currently in talks with the government to restore access as soon as possible. In the meantime, users can continue to utilize the company’s older AI models.
Existing Dispute with the Trump Administration
Tensions have been brewing between Anthropic and the Trump administration for the past few months. The conflict arose after the company refused to allow US security agencies to use its AI models for domestic surveillance and fully autonomous weapons systems. Consequently, the Pentagon placed Anthropic on a “supply-chain risk” list.
This dispute has escalated just as the company prepares for its IPO, with its market valuation projected to reach approximately $1 trillion (roughly ₹92 lakh crore).
Zoho Founder Sridhar Vembu: Globalization is Over
Until now, US efforts had focused primarily on preventing AI chips and semiconductor technology from reaching other nations—particularly China. However, this new ban on software has taken the world by surprise.
Sridhar Vembu, founder of the Indian tech company Zoho, expressed concern over this decision, describing it as a major wake-up call for India. Vembu declared that globalization is effectively over.
He argued that instead of relying on foreign companies, India must rapidly develop its own sovereign AI capabilities and promote open-source models, as any nation could cut off access to its technology at any time in the future.
