US blocks – India’s push for sovereign AI is colliding with sudden US restrictions on foreign access to two Anthropic models. Anthropic says the US government ordered it to block access to Fable 5 and Mythos 5 for foreign nationals on June 12, 2026, raising fresh pressure
On June 12, 2026, Anthropic disabled two AI models for foreign-national access. The move didn’t affect the company’s other models, but it landed like a warning shot for organizations in India that have been betting on US-developed frontier systems while trying to build sovereign capability at home.
The affected models were Fable 5 and Mythos 5. Anthropic said the US government ordered it to block foreign-national access to both models, and that it had to disable them.
For Indian technology leaders, the alarm isn’t only about what they can build. It’s about what happens when access to powerful systems changes “without warning. ” right when those models are already embedded into real operations—API-driven workflows. enterprise AI gateways. and production infrastructure.
The immediate spark for the scrutiny was Claude Mythos Preview. which Anthropic introduced on April 7. 2026 and described as a model with advanced cybersecurity capabilities. In May 2026, ORF reported that Mythos access had been limited to a small group of US enterprises. ORF also reported that India’s Finance Minister Nirmala Sitharaman called the challenge “unprecedented,” and that India was seeking access.
That sequence puts the tension in sharp focus: sovereign ambition is one track. but day-to-day continuity and compliance can hinge on a vendor’s ability to provide access from outside the US. For enterprises. public agencies. banks. and telecoms. the question becomes less abstract—how do they keep operations running when a frontier model can be suddenly switched off for them?.
The policy backdrop adds another layer of fragility. The Biden-era AI Diffusion Rule would have placed India outside the least-restricted group for advanced AI chip access. But the US Commerce Department rescinded that framework before it took effect in May 2025.
Access restrictions are also shifting from chips to the way technology is delivered. The House of Representatives passed the Remote Access Security Act on Jan. 12, 2026. If enacted, the act would broaden export controls to remote access through internet or cloud computing services.
This is where India’s internal gaps become more visible. India’s domestic foundation for sovereign AI is the IndiaAI Mission, approved in March 2024 with a ₹10,371.92 crore outlay. Its pillars include public AI compute. indigenous foundation models. datasets. startup financing. skills programs. applications. and safe and trusted AI tools.
But the missing pieces remain practical—and they map directly onto the risk exposed by sudden US model access limits. The policy base still leaves hard gaps in compute, model control, cloud capacity, datasets, security testing, and procurement rules for sensitive workloads.
Hardware is described as the hardest gap. Subsidies for GPU access and negotiations with chipmakers, cloud providers, and systems integrators can help, but India’s most advanced AI build-outs still depend heavily on global chip, networking, and software suppliers.
Open-weight models. the argument goes. can reduce dependence on proprietary US systems for many commercial and government use cases—especially where local languages and domain data matter. Yet for frontier cybersecurity, advanced coding, scientific work, and complex agentic tasks, India still has fewer proven domestic alternatives.
Sovereign AI also demands more than models. It requires data centers. reliable power. cooling. security controls. audit systems. and rules for which workloads must stay on India-hosted or India-controlled infrastructure. Across APAC. AI data center planning is increasingly described as a power. grid. and availability issue. not just a cloud-capacity issue.
ORF has also proposed a Trusted AI Corridor between India and the United States. It is not an agreement, but a negotiation path for model access under defined security conditions.
Next, the focus is on whether US lawmakers move RASA forward in the Senate, whether any follow-on India-U.S. technology agreement includes AI access protocols, and whether India can secure any chip, cloud, or model-access exception. For APAC technology leaders. model choice is already slipping into vendor-risk planning—because in the real world. access doesn’t just determine capability. It determines whether critical workflows keep running.
IndiaAI Mission sovereign AI Anthropic Fable 5 Mythos 5 Claude Mythos Preview ORF Nirmala Sitharaman Remote Access Security Act AI Diffusion Rule export controls model access APAC AI infrastructure Trusted AI Corridor