India must not depend on foreign AI models, as US export controls demonstrate potential access denial. The US government’s restrictions on Anthropic’s Fable 5 and Mythos 5 signal that advanced models may not remain globally available or continuously usable. This situation presents a significant AI sovereignty challenge for India, necessitating a shift towards self-sufficiency in artificial intelligence development.
The US employs export controls to slow the diffusion of strategic technologies, including encryption source code, advanced chips, and semiconductor equipment. The Anthropic restrictions offer an early glimpse of a future where frontier intelligence could become subject to such directives. Consequently, India and other nations cannot rely solely on partnerships with foundational AI tech companies, as these entities are not sovereign players.
India had previously requested access to Anthropic’s Mythos under Project Glasswing. This initiative aimed to understand the capabilities of frontier AI models for cybersecurity across various sectors, including banking and telecommunications. That access has now been disrupted due to the US restrictions.
US export control directives are not unpredictable. Historically, technology restrictions have shaped global information flows. Anthropic operates within the US state, meaning that if the White House decides a frontier model must be restricted, even the largest AI lab must comply. India is Anthropic’s second-largest consumer base, but being a customer does not equate to having control over access.
India has faced similar denials before, which historically led to a pivot in its own development pathway. For example, India’s nuclear program was built over decades, with strategic capabilities developed under the assumption that external access could not be guaranteed. However, the nature of AI development differs from nuclear technology; nuclear capabilities were anchored in physical infrastructure and reached strategic thresholds, while frontier AI development is still evolving.
To address this, India needs a continuity doctrine for AI, fostering multiple research labs and institutions rather than focusing on symbolic achievements. Learning from Singapore’s success in biotech, building foundational capabilities now is crucial. This approach aims to avoid perpetual reliance on imports and secure future agency in the rapidly advancing field of artificial intelligence.
The current situation leaves unresolved how India will bridge the immediate gap in access to advanced AI models. Future developments will likely involve increased domestic investment in AI research and infrastructure, potentially leading to new international collaborations that prioritize sovereign control over technology. The long-term implications for India’s technological independence and its role in the global AI landscape remain a key area to watch.