Claude logo, a keyboard, and a robotic hand in this illustration taken June 5, 2026. REUTERS/Dado Ruvic/Illustration | Photo Credit: Dado Ruvic

As the US has decided to block foreign nationals from accessing Anthropic’s Fable 5 and Mythos 5, Indian tech thought leaders have flagged concerns about the risks of using AI models developed abroad.

Though there is no direct impact yet, experts have urged Indian organisations and other users to treat this as a wake-up call as they transition to AI-led operations and deploy advanced large language model-based solutions. 

The US government on Friday issued an export control order, directing Anthropic to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the country. The list also includes foreign nationals who are working for the company. This has resulted in the abrupt disablement of the two models for the target customers. 

While announcing the abrupt suspension of access to Fable 5 and Mythos 5 models, Anthropic has apologised to its customers. “We believe this is a misunderstanding and are working to restore access as soon as possible,” it said.

Global concern

Though this move has not impacted access to all other Claude models, it has triggered a global uproar. Experts said that a possible abrupt suspension of AI services could seriously hurt the business operations.

Jaspreet Bindra, Founder of AI and Beyond, said the development was no surprise. “This was always a possibility. We have already seen this move before, a little bit with Microsoft suspending services to Nayara Energy. AI has emerged as a key utility like IT, water and energy,” he said.

“This is precisely why we should have our own sovereign AI. The bigger concern will be if this (blocking access) is extended to foundational models,” he said.

Zoho Chief Scientist Sridhar Vembu highlighted that globalisation is dead and that technology is the ultimate weapon. He asked Indian companies to adopt smaller open-source models from India and China and focus on cost-effective research and development.

Gundala Nagaraju, an AI governance expert and CEO of Astravion NextGen Catalyst Pvt Ltd, urged organisations to avoid overreliance on a single AI provider by implementing diversification and resilience strategies. “You should have a provider-agnostic technology ecosystem. You must assess possible business impacts and activate continuity measures to sustain critical operations. You should have a comprehensive visibility into AI dependencies,” he said.

Srikanth Velamakanni, Chairman of Nasscom, said that models would continue to improve over the next few months. “Managing this AI progress and making sure that AI is used responsibly requires coordinated global efforts. The arrival of Mythos just heightens the urgency around this need for global collaboration,” he said, pointing out at the US security concerns around bad actors getting access to Mythos.

Hemant Mohapatra, Vice-Chairman of LightSpeed India, said that beyond capital, building a foundational AI model that is competitive from scratch can cost anywhere between $250- $600 million.

“This is not as much about raising $50-60 billion, or raising it all at once. The early bottleneck is more on securing top talent and the GPUs,” he added.

Anthropic’s statement

The US-based AI firm has said that the government issued the export control order in response to objections from national security authorities. “They didn’t provide specific details of its national security concern. Our understanding is that the Government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5,” it said.

Anthropic contended that  it instituted strong safeguards that greatly reduced the risks of misuse of the model.

It, however, admitted that there is no perfect jailbreak resistance mechanism that is currently possible for any model provider.

“We aimed to make jailbreaks either narrow or very expensive to produce,” it said.

(Inputs from Rohan Das, Chennai)

Published on June 13, 2026