DeepSeek’s Success Rattles US Tech Market, But Positives Seen For India's AI Sector

DeepSeek, a Chinese AI startup led by Liang Wenfeng, has disrupted the global artificial intelligence landscape with its open-source model — DeepSeek-R1. While its success has rattled the US tech market, analysts see some key positives for India.

DeepSeek-R1, while rivalling models from giants like OpenAI and Meta Platforms, has been developed at a much lower cost of $6 million. The model’s focus on transparency and efficiency bypasses restrictions on advanced semiconductor exports to China, while its app has topped Apple’s App Store rankings.

"DeepSeek cracked something which everyone thought was expensive," Tanuj Bhojwani, former head at People+AI, told NDTV Profit. "They managed to create a machine that learns to think, reason, and backtrack."

CP Gurnani, co-founder of AIonOS and former chief executive officer of Tech Mahindra, drew parallels between DeepSeek’s approach and India’s advancements. "At Tech Mahindra, we developed our own AI in less than $1 million by using rented compute power and quality data, including rural contributions. We developed Project Indus which has Indian languages and its 40 dialects and we developed our own LLM," he said.

Gurnani said that what is being seen in the case of DeepSeek is more or less a similar path, and it reflects a broader trend in AI. "Open-source, cheaper hardware, and the theory of constraints" enabled a full-scale large language model at a much lower cost, Gurnani said, adding that this efficiency shows "you can do much more with less".

DeepSeek's success sparks optimism as the Chinese startup has been able to challenge American big-tech firms despite the restrictions imposed by Washington on the supply of advanced AI chips.

Ashok Chandak, president of the India Electronics and Semiconductor Association, highlighted that export restrictions on AI-related chips imposed by the US was seen as one of the constraints for growth in this sector. "There was chaos in Indian organisations about alternatives, but now people know we can achieve this,” he told NDTV Profit.

The cost reductions demonstrated by DeepSeek have profound implications for global AI dynamics. "Training costs are lower, and I don’t need Nvidia’s higher-strength GPUs. AMD or Qualcomm GPUs could suffice,” said Deepak Shenoy, CEO of Capitalmind. "This is driving fear in US markets, but it’s very good news for India in the long term."

Notably, India is making significant strides in the artificial intelligence space. The country is producing 16% of the world’s AI talent, and its AI market is expected to grow at 25-30% compounded annual growth rate by 2027, according to QS World Future Skills Index. Last year, the government approved a Rs 10,372-crore National AI Mission to boost the advancements in AI.

Indian GenAI startups saw a sixfold increase in funding in the September quarter of fiscal 2025, totalling $51 million, as per a report released by Nasscom. Platforms launched by TCS (WisdomNext), Wipro (Lab45), and Birlasoft (Cogito) showcase India’s push towards AI integration.

Bhojwani pointed out the need for greater data centre capacity to support such developments. "We still need significant investment," he said.

As AI costs drop and open-source models gain prominence, analysts believe India is well-positioned in the next phase of AI development. "If training costs come down, it’s a huge opportunity for India," Shenoy pointed out.