Exposing Indian Data to Foreign LLMs will Commoditise Our Knowledge, Says TATA Communications CEO AS Lakshminarayan

If Indian exposes data with foreign Large Language Models (LLMs)), it could lose control over its valuable knowledge and hence commoditise it, said AS Lakshminarayan, MD and CEO of Tata Communications. He stressed that India must build its own AI models—whether large language models (LLMs) or smaller versions—to retain control over its data.

“India should retain control over its own data, which is sovereign property. If we expose our data to external LLMs, the knowledge becomes commoditized,” Lakshminarayan told Outlook Business.

The USA recently tightened AI chip export controls, and with Trump potentially coming into power, the general fear is that such measures may increase. On being asked about the same, Lakshminarayan said, “It’s difficult to predict, and anything I say might be wrong because people are preparing for unpredictability. We’ll have to wait and see how things unfold.” Just to give some context, in January 2025, the US government imposed restrictions on exporting high-end AI chips to countries outside its key allies which includes Britain, Canada, Germany, Japan, South Korea and Taiwan. Now countries not considered as a closed ally including India, can import up to 50,000 advanced AI chips through 2027 without a license.

Also Read | JLR, Tata Communications Strengthen Partnership for Connected Car Tech

Lakshminarayan was speaking to Outlook Business at the launch of Tata Communications’ next generation cloud fabric named Tata Communications Vayu in Mumbai. The company claims that this cloud solution is built on over a decade of experience in cloud computing, covering single-product to multi-product solutions. To simplify, Cloud Fabric is an advanced cloud infrastructure that unifies multiple cloud platforms and computing resources into a single, cohesive network.

Vayu AI Cloud provides access to powerful NVIDIA GPUs on demand, making it easy to train and run AI models at scale. In October 2024, the company also tied up with chipmaker NVIDIA Hopper GPUs to enhance its AI cloud infrastructure. To simplify, NVIDIA Hopper GPUs are high-performance graphics processing units (GPUs) designed specifically for AI, data centers, and supercomputing.

The cloud platform is industry-agnostic, serving multiple sectors. However, it has extensive experience working with government agencies, BFSI (Banking, Financial Services, and Insurance), manufacturing, and retail, added Bhaskar Gorti, Executive Vice President, Cloud and Cybersecurity Services, Tata Communications.

Also Read | Lack of Enterprise Data in AI Models Both an Opportunity and Challenge: Tata Communications CEO

Speaking about the company’s collaboration with the government, Gorti told Outlook Business, “Tata Communications is engaged with various government entities, including the National Health Authority (NHA) and state governments, to support cloud infrastructure for large-scale projects. However, details on state-level engagements cannot be disclosed at this time.”

In Q3 FY25, Tata Communications reported a net profit of Rs 236 crore compared to Rs 45 crore in the last year period. The company's operating revenue increased to Rs 5,798 crore, compared to Rs 5,588 crore in the corresponding period last year.