World · South & Central Asia · 1 day ago
India faces risks in relying on US or Chinese AI models
The writer argues that India should be cautious about relying on leading AI models from either the United States or China.
In June, an American export directive led Anthropic to temporarily switch off access to two flagship models because it could not verify every user’s nationality.
Access returned on July 1, but the episode showed that US access could be withdrawn.
Chinese open models can be downloaded and adapted, but the writer says they may reflect Beijing’s political sensitivities.
For example, DeepSeek’s chatbot declined to say whether Arunachal Pradesh is an Indian state.
The writer warns against relying on either country for systems used in Indian banks, courts and public services.
India is backing domestic efforts, including planned trillion-parameter models by Sarvam AI and state-backed BharatGen.
The government is also reportedly considering up to 200 billion rupees for a new frontier AI fund, alongside funding approved for the IndiaAI Mission in 2024.
An AI researcher argues that India risks losing access to US models or importing Chinese political sensitivities if it relies on either country.
Anthropic switched off two flagship AI models on June 12 after a US export directive barred foreigners from using them.
Access to the models returned on July 1, but the episode showed that access could be withdrawn from Washington, the article says.
Chinese open models are described as capable, cheap and adaptable, but DeepSeek declined to say whether Arunachal Pradesh is an Indian state.
India is pursuing its own AI models and considering a frontier AI fund, alongside funding already approved for the IndiaAI Mission.
- Who
- India, whose AI development and public services are at issue.
- What
- An AI researcher warns against relying on US or Chinese models and urges India to build its own capacity.
- When
- The article was published on October 10, 2026; it also describes events on June 12, July 1 and in September.
- Where
- India; the events include a Brics summit in New Delhi.
- Why
- US access can be withdrawn, while Chinese models may reflect Beijing’s sensitivities.
This story does not have two clearly opposing sides.
be a pioneer in establishing a Brics AI open-source community
No country of 1.4 billion people should run its banks, courts and public services on either.
The IndiaAI Mission approved 103.7 billion rupees.
A US export directive barred foreigners from using Anthropic’s two flagship AI models, and the company switched them off.
Access to the Anthropic models returned.
At the Brics summit in New Delhi, Xi Jinping promised China would be a pioneer in establishing a Brics AI open-source community.
- Anthropic models
- Two flagship models were switched off on June 12 and access returned on July 1.
- IndiaAI Mission
- 103.7 billion rupees approved in 2024.
- Proposed frontier AI fund
- Government reportedly exploring an anchor investment of up to 200 billion rupees (US$2.08 billion).
- Sarvam AI
- Plans to build a trillion-parameter model.
- BharatGen
- State-backed; plans to build a trillion-parameter model.
Sources
Related news
China-India ties thaw, but military relations remain strained
Flash flood prompts evacuations along Nepal’s Budhi Gandaki River
Flood in central Nepal leaves five missing and sweeps away homes and bridges
TNSTC bus conductor arrested over alleged gutka transport
India condemns Vance’s remarks as US curbs green card filings at IT firms
Two killed in separate Bengaluru two-wheeler crashes
Two train drivers injured in separate stone-throwing incidents near Tuni and Samalkot
Humanitarian worker Steve Jones dies aged 77
India rejects UN rights office concerns over Delhi protest detentions
Sri Lanka president says India recognises its claim to Katchatheevu
Hyderabad probe links account network to nearly 450 cybercrime cases