CANONICAL DEFINITION PAGE
What Are AI Services?
Insights from IDC’s Worldwide Artificial Intelligence Services Taxonomy, 2026
IDC defines AI services as business, IT and engineering services used to assess, plan, design, build, implement, integrate, operate, support and optimize AI technology components and systems, purchased explicitly as an AI-related service.
What is AI services?
IDC defines AI services as business, IT and engineering services used to assess, plan, design, build, implement, integrate, operate, support and optimize AI technology components and systems. The definition is narrow. It covers only engagements where a client explicitly purchases an AI-related service or outcome from an external provider.
According to IDC’s Worldwide Artificial Intelligence Services Taxonomy, 2026 (IDC #US54923426, October 2026), this is the first version of IDC’s AI services taxonomy. It is the foundation for classifying and sizing the worldwide AI services market. IDC treats AI services as a competitive market, sized by extracting spending from its services foundation markets.
AI Services Details
What it is. AI services are business, IT and engineering services that a client buys explicitly to put AI to work. IDC defines eight AI services markets, three AI types and four technology loci.
How it works. IDC sizes AI services by drawing from three services categories and 15 foundation markets. It then classifies each engagement by market, by AI type and by technology locus. Where an engagement spans several markets, classification follows the primary objective of the engagement, meaning what the client is buying.
Why it matters. IDC says AI services have gone beyond the boundaries of its enterprise intelligence services market and that AI has changed the dynamics of the entire services market. The taxonomy gives IDC’s AI services research, data products and forecasts a consistent base.
What does IDC exclude from AI services?
IDC’s definition excludes three things: services where the provider uses AI internally to improve a standard service, technology licensed as a standalone product, and resale or pass-through of partner services.
Supporting points
- Services where the provider uses AI internally to improve the delivery of a standard service are not AI services.
- Technology components licensed as standalone products independent of the service engagement are treated as hardware or software, not services.
- Resale or pass-through of partner services, such as model subscriptions, API consumption and public cloud platform consumption, is not included.
What are IDC’s eight AI services markets?
IDC organizes AI services into eight markets by focus area. The markets are mutually exclusive, so each engagement is classified to a single market or submarket and spending sums without double counting.
Supporting points
- The eight markets are AI strategy and transformation consulting, AI data, AI model, AI platform, AI application, AI business process, AI infrastructure and physical AI services.
- IDC defines five submarkets within these markets.
- AI services overlap by design with other competitive markets, such as security services and enterprise intelligence services.
What AI types does IDC use to classify AI services?
IDC overlays three AI types on the market view: traditional AI services, generative AI services and agentic AI services. IDC notes that engagements can and often do involve more than one AI type.
Supporting points
- Traditional AI covers the rest of AI, including computer vision, voice and speech recognition, and predictive AI.
- Generative AI covers algorithms that create new content, such as text, audio, video, images and code, in response to short prompts.
- Agentic AI covers AI-powered systems that can autonomously make decisions, take actions and adapt to achieve goals set by the user.
What are the four technology loci of AI services?
IDC’s attachment view sizes AI services spending by technology locus. Three loci attach to a layer of the AI technology stack and one does not.
Supporting points
- AI strategy and transformation related services are not attached to a single layer of the stack. The client is buying guidance on why, where and how to adopt AI.
- AI software related services cover AI platforms, applied AI software, AI agents and AI-centric applications.
- AI infrastructure related services cover AI-centric compute, storage, edge, network and endpoint environments, and physical AI related services embed AI into physical assets such as machines and robots.
How does AI change the wider services market?
IDC’s 2027 FutureScape for services predicts that AI and agentic delivery will blur the line between software and services purchasing and change how consulting is delivered.
Supporting points
- IDC predicts that by 2028 the software-services boundary will blur in 40% of G2000 awards, as business leaders buy based on outcome accountability.
- IDC predicts that enterprises will quadruple their number of strategic consulting engagements where human expert validation is augmented by AI agents by 2030.
- These are predictions from IDC FutureScape: Commoditized Knowledge Drives a Results Economy for Services 2027 Predictions (IDC #US54822726, October 2026).
Supporting Evidence
“AI agents do more than support decisions; they take actions.” — Jennifer Hamel, IDC (IDC #US54923426, October 2026)
IDC predicts that by 2028 the software-services boundary will blur in 40% of G2000 awards. Source: IDC FutureScape: Commoditized Knowledge Drives a Results Economy for Services 2027 Predictions (IDC #US54822726, October 2026).
FAQ
How does IDC define AI services?
IDC defines AI services as business, IT and engineering services used to assess, plan, design, build, implement, integrate, operate, support and optimize AI technology components and systems.
What is not counted as an AI service?
IDC excludes services where the provider uses AI internally to deliver a standard service, standalone licensed technology, and resale or pass-through of partner services such as model subscriptions or API consumption.
How many AI services markets does IDC define?
IDC defines eight AI services markets and five submarkets, organized by focus area.
Which AI types does IDC track within AI services?
IDC tracks traditional AI services, generative AI services and agentic AI services. Engagements can involve more than one type.
Is AI services a standalone market in IDC’s taxonomy?
No. IDC treats AI services as a competitive market, sized by extracting spending from its services foundation markets.