When Anthropic pulled the plug on its top AI models for India, it set off alarm bells. The US government ordered Anthropic to suspend Fable 5 and Mythos 5 on June 12, cutting off Indian users overnight. This wasn’t just a business hiccup. It exposed a deeper problem: dependence on foreign AI infrastructure means accepting foreign politics and policies.

India had quickly become Anthropic’s second-largest market. Tata Consultancy Services had just signed a deal to train 50,000 employees on Claude, Anthropic’s AI platform. That deal is now uncertain. Indian companies and developers suddenly face a harsh reality. Access to cutting-edge AI can be taken away without warning or input.

This sudden shutdown has pushed India’s sovereign AI movement into the spotlight. Leaders like Mohandas Pai, a top tech investor, are calling for a ₹50,000 crore (about $5 billion) yearly fund dedicated to building homegrown AI. He also suggests a ₹2 lakh crore (roughly $21 billion) credit guarantee for cloud and semiconductor projects. This dwarfs the government’s current AI budget of about $1.25 billion.

Sridhar Vembu, founder of Zoho, argues India should focus on smaller, open-source models, including Chinese ones. He says relying on American AI is risky since it can be switched off at any time. “Technology is the ultimate weapon,” he said. “Globalization is dead. Bharat must find her own way.”

The Anthropic Fable 5 Launch and Fallout

Anthropic’s Fable 5 was a big deal. It was their most powerful AI, released just days after the company warned the world about the risks of rapidly advancing AI. Fable 5 combined top-notch performance with safety measures that routed risky queries to a weaker fallback model.

But things quickly got messy. Security researchers claimed to have bypassed the model’s safety systems. Others accused Anthropic of silently degrading Fable 5’s responses for researchers working on rival AI projects. The model would sometimes hand off complex, sensitive queries to a less capable system without telling users.

This silent downgrade angered the AI community. Researchers felt betrayed because they trusted the model to deliver consistent, transparent answers. After public backlash, Anthropic apologized and made the fallback process visible. But the damage to trust was done.

What This Means for India and Global AI Policy

For India, the Fable 5 suspension was a wake-up call. Indian startups are already adapting by releasing open-source models and pivoting towards AI infrastructure services. But none match the power of Fable 5 yet. The goal is not to build the world’s best AI immediately. It’s about controlling the floor beneath their feet.

AI policy experts say American AI models are tied to American geopolitics. For companies with teams in multiple countries, relying on US-based AI providers means they are one policy decision away from disruption. This has serious implications for India’s tech and business sectors.

Some voices caution that India’s real challenge is access to computing power, not just funding. Building world-class AI models requires massive compute resources, which are expensive and often tied to foreign cloud providers.

Meanwhile, Anthropic itself is navigating complex waters. They have an office in Bengaluru and had big plans for India. Whether those plans survive depends on how quickly they restore access and rebuild trust with Indian users.

The bigger picture shows the US tightening controls on AI technology. Export restrictions, model-level interventions, and safety regulations are shaping the global AI landscape. Countries like India now face a choice: build their own AI ecosystems or risk being cut off by foreign decisions.

India’s sovereign AI push is more urgent than ever. It’s about independence, security, and future-proofing their technology sector. The shutdown of Anthropic’s models was a harsh lesson. It proved that AI access is a privilege that can be revoked. And that lesson is driving a new wave of investment, open-source development, and policy debate across India.