Telco AI shifts from strategy to execution

According to GSMA Intelligence, 2025 is the year telcos are putting their money where their mouth is when it comes to AI.

The recent ‘Telco AI: State of the Market, Q2 2025’ report suggests the industry has shifted from AI strategy meetings to real-world action.

The first order of business has been a practical, heads-down focus on the bottom line. As the report puts it, if 2024 was for establishing the “strategic rationale for AI, 2025 is about assessing progress.”

First, the easy wins

Right now, operators are tackling the low-hanging fruit, and the results are most obvious in customer support. AI-powered chatbots and automated call centres make up a massive 47% of all tracked AI projects.

The main goal is simple: use automation to achieve cost savings. Networks are the second most common area for AI, attracting around 20% of deployments for tasks that keep things running smoothly, like preemptive maintenance.

80-85% of these telco AI projects are designed to cut internal costs and improve efficiency. Driving new revenue is a much smaller piece of the pie, representing only 10-20% of deployments so far.

And these aren’t just pilot schemes. A solid 60% of these AI systems are already live and part of the day-to-day business. The other 40% are in trial or planning stages. GSMA Intelligence believes this is the new normal—a constant cycle of testing, deploying, and innovating with AI.

Edge computing is the next frontier for telco AI

But while saving money is today’s priority, the report suggests the real story is what’s happening at the network’s edge. This is where the industry is placing its bets for future growth and new revenue.

The heavy-duty work of training AI models still happens in massive cloud data centres. But using that trained AI in real-time – a process called ‘inference’ – is increasingly being moved to the edge of the network, closer to the user. The benefits are huge: it can be 30-40% cheaper, reduces lag, and keeps sensitive data more secure.

This is opening the door to a new wave of services for business clients, with early use cases focused on things like industrial robotics, smart security cameras, and complex ‘digital twins’ for factory floors.

However, the main challenge, the report notes, isn’t the tech itself but figuring out how to make money in a market that “is quickly becoming a crowded field.”

A new cast of AI partners for telcos

This AI revolution isn’t a solo act. A whole ecosystem of tech partners is involved, but a few names are dominating the headlines. In the all-important GPU market, Nvidia has a “significant first-mover advantage.”

More surprising, however, is the rise of Perplexity, an AI search engine that has become a key partner for many telcos. The report speculates this could be a move intended to hedge “against Google’s dominance.”

Looking ahead, this AI push is forcing operators to develop new muscles. Upskilling their own staff is now a top priority for 63% of operators, who know they need the in-house talent to manage this new technology.

Telcos are clearly using AI as a practical tool to solve today’s pressing cost issues. At the same time, they’re carefully laying the groundwork at the edge for a future where AI is the main engine for growth. The next few years promise a period of innovation as the industry learns how to maximise the potential of AI.

(Photo by Leon Kohle)

See also: SK Telecom releases compact AI model for mobile use on Hugging Face

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Tags: 5G, connectivity, gsma, mobile, networks, Operators, research, study, telecoms