Telco AI deployments increasingly focused on monetisation – report

• 35% of recent telco AI deployment plans include a revenue objective, according to GSMA Intelligence
• Sovereignty trends are strengthening the role that telcos play in AI
• The GSMAi team highlights particular revenue-generating opportunities in GPUaaS and AI factories

The telco sector is shifting towards a new age of AI monetisation, with more than a third of AI deployments announced in the past six months coming with a revenue objective, according to a report from GSMA Intelligence (GSMAi).

In its latest Telco AI: State of the Market report (subscription required), GSMAi found that 35% of telco AI deployments announced during the past six months included direct revenue objectives, including the provision of AI inference applications and GPU-as-a-service options. This signals how telcos are increasingly thinking about more than just driving efficiencies from their AI deployments. 

In the report, the GSMAi team splits monetisation into three categories: Strong opportunities, such as GPUaaS, sovereign AI, and vertical solutions; medium viability, such as managed AI and enhanced enterprise bundles; and challenging propositions, such as foundation model training, and open AI marketplaces.

The report particularly highlights the increasing role in telco strategies of AI factories and the potential revenue uplift these provide for operators. AI factories are specialised, high-performance data infrastructure facilities that turn raw data and massive computing power into automated intelligence, predictions, and digital agents. According to the report, AI factories carry a potential revenue uplift of 5% for mobile operators, and offer telcos a direct route into the  growing AI infrastructure market.

According to the report’s authors: “Capturing even a small share of enterprise AI infrastructure demand can generate meaningful new datacentre revenues. GSMA Intelligence modelling indicates that even a modest 5–10% conversion rate of cloud revenues translates into an uplift of around 1.5–3.0 pp [percentage points] for service revenue growth. This extends to around 5 pp at a 20% conversion rate.”

Sovereign AI is another big opportunity and, according to the GSMAi, it has now “gone mainstream” across all regions. While it is most prevalent in the US and Canada, where around 40% of operators have launched a product: Even in Latin America (the least prevalent region) 20% of operators have launched some form of commercial sovereign AI product.

But what is driving the demand for sovereign AI services? The GSMAi figures show operators are seeing the greatest demand from more heavily-regulated industries, such as finance, healthcare and defence, especially in Europe and Asia Pacific, while regulatory mandates, such as data residency requirements, are the leading drivers in the Middle East and Africa and Latin America. 

While this all sounds promising, the report does sound one particular note of caution with regards to investment scale, as it contextualises telco levels of AI infrastructure capex compared with that of the hyperscalers. According to the GSMAi, hyperscalers continue to invest in AI-related capex “at an unprecedented pace”. In 2025, for example, Amazon invested $92.3bn, Google invested $54.4bn, and Microsoft invested $53bn. 

By comparison, during the same period, the main telcos in China invested the most in AI infrastructure last year with a combined $19.6bn, while telcos in Germany and the UK invested $3.6bn per country: The numbers highlight not only the differences in strategies and economic muscle, but also how challenging it will be for telcos to compete with the hyperscalers.

- James Pearce, Editor, TelecomTV