Why telcos have the most to gain from AI

As the industry scales into enterprise-wide deployment, telco leaders are increasingly focused on building in-house AI to deploy and govern the technology at scale. A report by TCS revealed that nearly half (48%) of telco operators worldwide have started enterprise-wide deployment of AI, with as many as 81% integrating AI into their networks and operations.

Akhilesh Tiwari, president – communications, media and information services at TCS, sat down with Capacity and explained how the company is partnering with global telecom leaders to navigate the next chapter of digital transformation.

“AI is not yet fully deterministic to the extent that telcos want it to be, but the pace at which it’s moving forward is remarkable,” he explained. “Optimisation of the network and optimisation of spectrum – their most expensive asset – are areas where we see tremendous scope for AI in the coming years.”

The agentic tipping point

Industry leaders continue to say 2026 is a ‘tipping point’ for agentic AI, as the technology moves from the pilot and exploration stage to the operation and deployment stage. Here, the telecoms industry is somewhere AI can make a real difference, but Tiwari explained that skills gaps, among other challenges, remain a significant barrier to innovation.

“Execution is lacking right now, but less than 5% cited budgets as an issue,” he said, listing legacy infrastructure and data quality as critical problems that hinder companies. “Telcos are very technology-centric companies by nature, but the skills profile of their workforce is a direct corollary of that. That’s where the role of a partner becomes important.”

According to Tiwari, most TCS customers and operators are seeking help on AI strategy and almost none on AI operations. He said this is how TCS can assist companies, by creating demos and helping them put things into production.

“When we create that pathway, the rest of their organisation follows along and skills enablement can happen,” he said. “We faced a similar challenge ourselves. We started a reskilling programme last year and 250,000 employees participated.”

Finding out 81% of telcos are deploying AI in their networks and operations was a surprising finding for TCS, Tiwari said, as they understand telcos now prioritise customer experience. However, this cannot be delivered unless the company has strong integration across its BSS, OSS and other systems.

“If the network doesn’t provide a reliable service, then no matter what you put at the top of your customer experience stack in terms of agentic AI, you won’t be able to deliver a consistent and reliable experience,” Tiwari explained. “Network and operations also have a tremendous amount of data in a standardised format.”

Creating ‘meaningful’ AI

For Tiwari, 2026 will be the year the industry will be where agentic AI becomes tangible. However, while the technology has entered the mainstream this year, the TCS report finds that adoption is still lagging.

“The number one reason is, in my opinion, the legacy nature of their application infrastructure,” Tiwari said. “To deploy meaningful agentic AI, those systems have to be able to act across the full flow – from order to fulfilment, to billing, to management – and currently, they don’t allow for that kind of real-time interaction.

“Unless you have seamless, real-time access across those systems, agentic AI will struggle to deliver on its promise.”

This isn’t just a telco-related challenge, Tiwari explained, but extends across all large corporations with multiple layers of technology and IT applications. This is where TCS comes in, to help modernise legacy applications.

“We’ve taken ownership on an outcome basis: we will deliver your end-state architecture, jointly defined, within a set time period and a fixed cost,” Tiwari said. “That creates a fantastic foundation for seamless agentic AI implementation.”

TCS is operating in an industry environment where the maturity of agentic AI technology improving rapidly. Yet part of their role involves managing enterprise anxieties about the technology.

“The progress made just in the last six weeks is something I would not have anticipated six months ago,” Tiwari said. “Telco is a national critical infrastructure, so operators need to be certain over their agents – they can’t have agents do something that brings down an entire network.”

Another big consideration for companies is return on investment (ROI), particularly if they’re planning on investing heavily in new technology. Tiwari advised that it’s important for a business to focus on use cases that are going to deliver customer value to build shareholder trust, rather than launching large programmes.

“Telcos are investing heavily in new technology, but can be slow in switching off older systems, which leads to higher costs and slower systems,” he said.

Likewise, being unable to retire older technology means, in Tiwari’s mind, that ROI in 5G and associated IT spend hasn’t been felt yet.

“Getting to whatever the latest investment actually is, and retiring what came before it, should be priority number one – that’s how they can demonstrate to shareholders that new technology investment will generate return,” he explained.

TCS: From reliability to scale

For TCS, its competitive edge is its execution. Tiwari explained that the company continues to pride itself on being a consistently reliable partner.

“That is our DNA and how we retain business,” he said. “A lot has been done on AI strategy across the industry, but very little on AI operations.

“We’re signing up to deliver a specific reduction in customer churn, for example. If that doesn’t happen, the customer doesn’t pay – which increases their confidence because they know we have skin in the game.”

To modernise successfully, TCS is harnessing its AI knowledge to help its partners innovate with speed. The company can deploy AI agents in the development lifecycle to quickly assess where the biggest legacy vulnerabilities are, remediate and test them at a speed that wasn’t possible six months ago, Tiwari explained.

“We have to deliver modernisation continuously to ensure current networks continue to run,” he said.

He added that the company has also formed relationships with all the major foundational labs, including a significant partnership with OpenAI to make their products available to the entire Tata Group. As part of the alliance, OpenAI became the first customer of TCS’s HyperVault data centre business, which will commence with 100 megawatts (MW) of capacity and with potential to scale to one gigawatt over time.

“This gives us what we call an infrastructure-to-intelligence layer,” Tiwari explained, referring to the company’s multiple high-end partnerships. “This is an end-to-end capability that I don’t think anyone else in the industry can match, from owning the infrastructure layer right through to delivering intelligence to our customers.”

Looking ahead, TCS is eager to grow its position as a world leader in AI. So far, Tiwari explained that market momentum is helping them get there.

“We’re seeing something very concrete; customers who previously had fixed budget for a body of work are finding they can now accomplish their work for significantly less,” he said. “There’s concern about job losses, but we simply don’t see that – we see customers doing more with the budgets they have.”

Both internally and externally, he explained how TCS is helping to generate excitement about AI and its future.

“TCS is in a very good place currently,” he added. “The partnerships we’ve established and our focus on making AI real and outcome-focused – that excites both our employees and our customers.

“I’m quite bullish, particularly for telcos, because they probably have the most to gain from this out of any industry.”

This article forms part of the latest edition of Capacity Magazine, which you can read HERE.

Cisco’s Jeetu Patel: Telcos can’t ‘stay on the sidelines’ with AI

How AI could put voice back at the centre of telco revenue