From connectivity provider to intelligent, AI-native operator — our research, told in six chapters.
Start hereChapter 1 · The pressure
Ch.1ReportTelecom industry trends
Operator shareholder returns have cleared the cost of equity for the first time in years, but connectivity revenue is still growing at low single digits and capex is past its peak. The gap between leaders and the rest is widening, and scale rather than spend is becoming the strategic variable.
Then read
Ch.4Executive briefingTelecom industry trends
Fraud losses on operator networks are still rising, but the bigger shift is regulatory. Operators are being asked to block, verify and share on behalf of the whole payments ecosystem, and to prove that they did.
Ch.3ReportTransformation
Operators sit on some of the richest data in any industry, yet data is the barrier executives most often cite to scaling AI agents. The fix is less about new platforms and more about governed data products, a shared semantic layer, lineage and consent built in from the start.
Our research follows one argument in six chapters: why the operating model is under pressure, where data and AI create value, what has to be built first, the proof it works, who does the work, and how to get there.
Sourced, public figures in chapter order — from the pressure on returns to the rules for AI. Each links to the insight that puts it in context.
3 insights
Ch.1ArticleFuture of telecom
Satellites that connect to ordinary phones have moved from demonstration to service, and regulators are clearing the spectrum path. The strategic question for operators is who controls the satellite layer, and who captures the value.
Ch.4Executive briefingTelecom industry trends
Fraud losses on operator networks are still rising, but the bigger shift is regulatory. Operators are being asked to block, verify and share on behalf of the whole payments ecosystem, and to prove that they did.
Ch.2ArticleAI in telecom
Operators now have credible evidence that GenAI and agents can cut cost per contact and lift resolution. Customers remain wary, regulators are watching for AI gatekeeping, and the programmes that work treat the handoff to a person as a feature, not a failure.
Ch.1ReportTelecom industry trends
Operator shareholder returns have cleared the cost of equity for the first time in years, but connectivity revenue is still growing at low single digits and capex is past its peak. The gap between leaders and the rest is widening, and scale rather than spend is becoming the strategic variable.
Ch.5Point of viewFuture of telecom
Operators are declaring Level 4 autonomy in individual domains and setting group-wide targets for 2028–2030. The evidence shows that the hard part is not fault management but change, and that agents only work on top of trusted data.
Ch.5Point of viewAI in telecom
The EU has pushed high-risk AI obligations back to December 2027, but transparency rules already apply and autonomous networks are arriving faster than the rulebook. Operators need governance that works at network speed, with graded autonomy and a stop button that really stops.
Ch.5Point of viewFuture of telecom
Operators run more AI pilots than almost any other industry, yet few have agents executing multi-step work at scale. The gap is not the models. It is data, workflow redesign and supervision.
Ch.2ArticleAI in telecom
Most operators can predict who will leave. Far fewer can decide what to do about it, for whom, and at what cost. The value sits in the decision layer that links propensity, lifetime value and a governed next-best-action engine.
Ch.3Executive briefingTransformation
Operators have spent years trapped between costly legacy stacks and transformation programmes that overrun. Open APIs, ODA components and AI-assisted code migration make an incremental path credible, provided data and decommissioning are treated as first-class goals.
Ch.2ArticleTelecom industry trends
5G subscriptions have passed three billion, but the revenue has not followed. Standalone cores, network slicing, fixed wireless access and network APIs are where the business case now rests, and each demands a different commercial muscle.
Ch.4ArticleTransformation
Industry surveys have put telecom revenue leakage at around 1.5% of revenue for years, and fraud adds more on top. The fix is not another dashboard but full-population controls, clear ownership and AI that works on reconciled data.
Ch.3ReportTransformation
Operators sit on some of the richest data in any industry, yet data is the barrier executives most often cite to scaling AI agents. The fix is less about new platforms and more about governed data products, a shared semantic layer, lineage and consent built in from the start.
Try a broader term, or browse all insights.
Growth, pricing, fixed-mobile convergence and the forces reshaping operators.
From network AI to supervised agents in care, fraud and revenue assurance — and how to govern them.
BSS/OSS data foundations, platform modernisation and operating-model change.
What the operator of 2030 looks like — and how to get there.
Each chapter pairs our research with the page on this site where the idea is put to work.
Flat ARPU, heavy 5G and fibre capex, fraud and scams, churn, legacy BSS/OSS and new AI rules.
Ch.1ArticleFuture of telecom
Satellites that connect to ordinary phones have moved from demonstration to service, and regulators are clearing the spectrum path. The strategic question for operators is who controls the satellite layer, and who captures the value.
Ch.1ReportTelecom industry trends
Operator shareholder returns have cleared the cost of equity for the first time in years, but connectivity revenue is still growing at low single digits and capex is past its peak. The gap between leaders and the rest is widening, and scale rather than spend is becoming the strategic variable.
Eight value-chain domains, from network build to the regulator — and where data and AI pay back in each.
Ch.2ArticleAI in telecom
Operators now have credible evidence that GenAI and agents can cut cost per contact and lift resolution. Customers remain wary, regulators are watching for AI gatekeeping, and the programmes that work treat the handoff to a person as a feature, not a failure.
Ch.2ArticleAI in telecom
Most operators can predict who will leave. Far fewer can decide what to do about it, for whom, and at what cost. The value sits in the decision layer that links propensity, lifetime value and a governed next-best-action engine.
Ch.2ArticleTelecom industry trends
5G subscriptions have passed three billion, but the revenue has not followed. Standalone cores, network slicing, fixed wireless access and network APIs are where the business case now rests, and each demands a different commercial muscle.
A governed telecom data platform and a 12-layer target architecture, built once and reused.
Ch.3Executive briefingTransformation
Operators have spent years trapped between costly legacy stacks and transformation programmes that overrun. Open APIs, ODA components and AI-assisted code migration make an incremental path credible, provided data and decommissioning are treated as first-class goals.
Ch.3ReportTransformation
Operators sit on some of the richest data in any industry, yet data is the barrier executives most often cite to scaling AI agents. The fix is less about new platforms and more about governed data products, a shared semantic layer, lineage and consent built in from the start.
Seven working accelerators — led by the SENTINEL fraud and revenue-assurance command centre.
Ch.4Executive briefingTelecom industry trends
Fraud losses on operator networks are still rising, but the bigger shift is regulatory. Operators are being asked to block, verify and share on behalf of the whole payments ecosystem, and to prove that they did.
Ch.4ArticleTransformation
Industry surveys have put telecom revenue leakage at around 1.5% of revenue for years, and fraud adds more on top. The fix is not another dashboard but full-population controls, clear ownership and AI that works on reconciled data.
Supervised AI agents doing routine work end to end, inside guardrails people set and audit.
Ch.5Point of viewFuture of telecom
Operators are declaring Level 4 autonomy in individual domains and setting group-wide targets for 2028–2030. The evidence shows that the hard part is not fault management but change, and that agents only work on top of trusted data.
Ch.5Point of viewAI in telecom
The EU has pushed high-risk AI obligations back to December 2027, but transparency rules already apply and autonomous networks are arriving faster than the rulebook. Operators need governance that works at network speed, with graded autonomy and a stop button that really stops.
Ch.5Point of viewFuture of telecom
Operators run more AI pilots than almost any other industry, yet few have agents executing multi-step work at scale. The gap is not the models. It is data, workflow redesign and supervision.
Know your stage with evidence, then move in phase-gated steps from pilot to scale.
Our research informs the tools we build with operators. Put the ideas to work.
Benchmark data, AI and operating-model maturity across the dimensions that separate leaders from the pack.
Take the maturity assessmentSee AI agents working real telecom cases — with owners, autonomy levels, approvals and live metrics.
Meet the digital workforceThe data, AI and governance layers an operator needs to move from pilots to production at scale.
Explore the frameworkNew perspectives on AI, data and transformation in telecom — a few times a month, no noise.