AI contributes about 2.5% of telecom operators’ revenue, according to GSMA Intelligence’s third-quarter report on the state of the AI market in telecoms, as reported by SDxCentral on 2 November. SDxCentral says that compares with a capex envelope equivalent to 15-20% of profits. The same report ranks Nvidia first among publicly announced telecom AI partnerships and calls it “the sun of the AI ecosystem”.
The report is built from publicly announced AI deployments across more than 760 operators worldwide. About 66% of the deployments it tracks are now live, up from around 60% the previous quarter. Cost-saving projects still outnumber revenue projects 62% to 38%, about 1.6 to 1 by our arithmetic.
The key insight: GSMA Intelligence’s Q3 2026 telecom AI tracker, as reported by SDxCentral, puts AI at about 2.5% of operator revenue, while two thirds of tracked deployments are live and cost-saving projects outnumber revenue projects 62% to 38%, about 1.6 to 1 by our arithmetic.
GSMA Intelligence’s own summary says the Q3 2026 edition of its Telco AI: State of the Market series shows “two thirds of AI deployments tracked across the telecoms sector now live”. SDxCentral, which reported the findings, gives the full split: 66% of tracked initiatives are live, 4% remain in trials and 30% are classified as commitments. Operators recorded an average of 1.6 publicly announced deployments each.
The mix of goals barely moved. SDxCentral reports that the split between cost-saving and revenue objectives went from 61% and 39% in Q2 to 62% and 38% in Q3. Customer care was the largest category, at 43% of tracked deployments in Q3, down from 44% in Q2; sales and marketing accounted for 27%, and network applications for 17%.
GSMA noted potential space for “significant headroom” in the business-to-business market, which accounts for 25-30% of telecom revenue overall, according to SDxCentral. It also stressed, the outlet reports, that announcing AI services or increasing deployment numbers does not necessarily translate into incremental revenue or sustainable returns.
Share of telecom operators reporting each revenue-generating AI use case in Q3 2026. Source: GSMA Intelligence, Telco AI: State of the Market Q3 2026, as reported by SDxCentral, 2 Nov 2026.
What Operators Sell
AI-enhanced connectivity was the most widely offered revenue-generating AI use case, reported by 75% of operators. GPU-as-a-service followed at 69%, edge and distributed inference at 57%, industry-specific AI solutions at 54% and sovereign computing at 37%, per SDxCentral’s account of the report.
Infrastructure, meaning GPUs, data centers, dark fiber and related physical assets, represents approximately 25% of telecom AI spending, the report found. Smaller operators spend a greater share of their revenue on AI, at 6.4%, especially in the Middle East, while bigger operators are scaling deployments beyond the pilot phase.
Business Pill · Capex and Depreciation
A FourWeekMBA explainer on how capital spending turns into depreciation on the income statement. It teaches the general idea only and says nothing about any company in this story.
Who Supplies Telecom AI
On the supply side, Nvidia ranks first among publicly announced telecom AI partnerships for both large and smaller operators, SDxCentral reports, which is why GSMA billed it as “the sun of the AI ecosystem”. AMD ranks second and Intel third among GPU suppliers, with that order reversed among smaller operators.
Amazon Web Services leads the hyperscaler partnership rankings for operators of all sizes, ahead of Microsoft and Google. Among frontier AI labs, Perplexity came first, ahead of OpenAI and DeepSeek; GSMA Intelligence put that lead down partly to Perplexity’s scope for customization within telecom-specific propositions, according to SDxCentral.
The report’s researchers concluded: “Deploying too slowly risks missing productivity and revenue opportunities. Deploying too quickly risks compromising trusted infrastructure.”
As we read SDxCentral’s account, most of the operators’ tracked AI deployments are still aimed at their own costs: customer care alone is 43% of tracked deployments, and the cost-saving share edged up from 61% to 62% in a quarter.
As we read the supplier rankings, the companies operators name most often as partners sit below them in the stack: Nvidia first among chip suppliers, Amazon Web Services first among hyperscalers.
GSMA Intelligence researchers, as reported by SDxCentral, 2 November 2026
“Deploying too slowly risks missing productivity and revenue opportunities. Deploying too quickly risks compromising trusted infrastructure”
Three Implications
FOR TELECOM OPERATORS GSMA Intelligence found AI contributes about 2.5% of operator revenue and pointed to potential headroom in business-to-business sales, which account for 25-30% of telecom revenue, according to SDxCentral.
FOR CHIP SUPPLIERS Nvidia ranks first among publicly announced telecom AI partnerships for both large and smaller operators, with AMD second and Intel third among GPU suppliers, an order reversed among smaller operators, SDxCentral reports.
FOR SMALLER OPERATORS The report found smaller operators spend 6.4% of their revenue on AI, especially in the Middle East, while bigger operators are scaling beyond the pilot phase.
The framework reads: “and at any given moment, the binding constraint sits at a different layer than it did six months ago.”
As we read it, GSMA’s telecom tracker shows operators buying AI from the layers below them, chips and hyperscale clouds, while AI is still about 2.5% of their own revenue. The framework is the owner’s argument and does not discuss telecom operators.
The full GSMA Intelligence report is gated; the figures here are as reported by SDxCentral, alongside GSMA’s own public summary. SDxCentral’s headline speaks of return on investment below 3%, but the figure in its text is AI’s share of operator revenue, about 2.5%; we found no separate ROI figure. The sources we read do not say how the capex envelope of 15-20% of profits is defined or which operators it covers.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
GSMA Intelligence’s Q3 tracker, as reported by SDxCentral, puts AI at about 2.5% of telecom operator revenue. Two thirds of tracked deployments are now live, but cost-saving projects still outnumber revenue projects 62% to 38%, and Nvidia ranks first among the sector’s AI partners.
A note on sourcing. The figures come from GSMA Intelligence’s Telco AI: State of the Market Q3 2026 report as reported by SDxCentral on 2 November 2026, alongside GSMA Intelligence’s public summary of the report. The full report is gated. The 1.6-to-1 ratio is our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
