企业AI：电信运营商的最佳机遇是什么？

Executive Briefing ENTERPRISE AI: WHAT ARETHE BEST OPPORTUNITIESFOR TELCOS? This report examines the key opportunities for telecom operators tomonetise enterprise AI and secure long-term relevance within theemerging AI ecosystem. Foreword Methodology This report presents insights from a recent research programme, comprising: •In-depth interviews with fourteen senior decision-makers across communication serviceproviders (CSPs), wholesale connectivity infrastructure providers and data centre operators. •Global enterprise survey (August 2025):Targeted survey ofmore than303enterprise AIdecision-makers, primarily from organisations with >1,000 employees across North America,Europe, the Middle East and Central Asia, covering industry verticals including retail, financialservices, professional services,manufacturing andhealthcare.•Quantitative forecast of AI-driven traffic on telecom networks (2025–2028), informed byinterview insights, survey findings, secondary research and STL Partners’ proprietary datasets. The aim was to understand how AI is impacting the telecom industry and to identify key monetisationopportunitiesfor operators. This isthe firstpieceof a two-part series,whichfocuses on monetisationopportunities, whileparttwowill provide practical recommendations for addressing them. Editorial independence This report has been prepared by independent consulting and researchcompanySTL Partners andwas commissioned by Cisco. STL Partners maintainseditorial independence. Mentions or allusions to companies or products inthis document are intended as illustrations of market evolution and are not included as endorsementsor product/service recommendations. Executivesummary The systematic and widescale adoption of AI technologiesis creating opportunities for telcos tosupport and enable enterprise customers in their AIjourneys.Broadly,STL Partners splits telcos’enterpriseAIbusiness opportunities into two buckets:AI infrastructure that enables enterprises’deployment of AI,and AI-based software applications and services offered by telcos to their enterprisecustomers. AI infrastructure: Connectivity and compute Overall, AI adoption is unlikely to transform business broadband and cloud connectivity requirementsin thenext 2–3 years. The immediate connectivity opportunities for telcos lie in high-performance,specialised connectivity and compute services rather than significant internet and cloud connectionupgrades.These include: Enterprise connectivity(around 10% of a standard,full-service telco’s annual revenue):Connectivityremains telcos’ core business, so it is the natural starting point. Security-sensitive sectors, includingfinancial services, defence and pharmaceuticals, are expandingon-premise dedicateddata centreandthird-party datacentrecapacity to build and run AI models in controlled environments, securingsensitiveproprietary data. These facilities depend on secure, high-reliability, high-performanceconnectivity, a requirement that telcosshould seek to address. Data centre interconnect(DCI):Explosive growth in hyperscale, neocloud and regional facilities,accelerated by agentic AI systems generating constant east-west traffic, is driving demand fordiverseinterconnectlinks between datacentres:bothphysical and supplier diversity.Regional and neocloudmarkets represent an accessible entry point for telcos seeking to capture this growth. AI factories:AI factories arelarge data centre facilitiespurpose-built for training, fine-tuning anddeployment ofAI models,ranging fromtensto hundreds of megawattsinpowercapacity.AsSTLPartners’researchhaspreviouslyexplored,telcoshave already moved into this space–andwe believethis is an opportunity specifically for telcos in marketswithlimited hyperscaler presence or marketswith politically-driven requirements for more data sovereignty. Edge AI:Wholly cloud-based AI processingmay belimited bypowergrid access,data privacy andlatencyconstraints. In some geographies (e.g.,around Dublin),the availability of power and timelyaccess to the electricity grid are curtailing the buildout of larger AI factories.Recent hardwaredevelopments have made it possible for AIinferenceworkloads to run on smallerphysical footprints.Together, these advances reassertsmaller‘edge’ data centresas a practical,local,compliant and cost-efficient place to runsomeAIworkloads. Services:AIsoftwaresolutions, managed services andadjacent offerings AI solutions:Telcos can support enterprise AI adoption by acting as neutral, vendor-agnosticplatforms forexploration, evaluationandsecurescaleddeployment, helping organisations avoid lock-in while meeting security and compliance needs.Telcos shouldembed AI into their own B2B offerings,to improve customer experience and strengthen proposition stickiness. Finally, they can develop anddeliver AI tools directly-from horizontal solutionssuch aschatbots and agents to vertical, industry-specific applications, creating differentiated value beyond connectivity. Cybers