CANONICAL DEFINITION PAGE
What Is Artificial Intelligence?
Insights from IDC’s Worldwide AI and Generative AI Spending Guide Taxonomy, 2026
IDC defines artificial intelligence as systems that learn, reason and self-correct, and it groups AI into three types: traditional AI, generative AI and agentic AI.
What is artificial intelligence?
Artificial intelligence is a class of systems that learn, reason and self-correct. According to IDC, these systems hypothesize and formulate possible answers from available evidence, are trained by ingesting vast amounts of content, and automatically adapt and learn from their mistakes and failures. This matters because IDC organizes its AI research, forecasts and spending guides around this definition and around three AI types, so the term means the same thing across IDC work.
According to IDC’s Worldwide AI and Generative AI Spending Guide Taxonomy, 2026: Release V2 (IDC #US54883926), the taxonomy serves as the framework for how IDC organizes its AI and generative AI research and forecasts. IDC research director Karen Massey calls AI “one of the most transformational technologies of our time.” IDC’s Artificial Intelligence Core Software Taxonomy, 2025 (IDC #US53964025) splits AI into traditional, generative and agentic categories.
Artificial Intelligence Details
What it is. Artificial intelligence, as IDC defines it, covers systems that learn, reason and self-correct. IDC’s taxonomies then sort AI offerings into traditional, generative and agentic categories, so one definition supports research, forecasts and spending guides.
How it works. IDC views AI through three categories. Traditional AI covers the tools used by data scientists and machine learning developers from experimentation to production. Generative AI covers systems that learn from training data to synthesize new content such as text, images, audio, video and code. Agentic AI covers systems that make decisions and take actions toward goals without constant prompting. IDC then maps products, spending and use cases onto these categories.
Why it matters. A shared definition keeps market sizing and vendor comparisons consistent. IDC’s AI and Generative AI Spending Guide provides forecasts by industry, by country and across more than 40 use cases, which lets buyers and suppliers place their investments within one framework.
What are the three types of AI in IDC’s framework?
IDC separates AI into traditional AI, generative AI and agentic AI. The types differ in what they do with inputs and how much autonomy they have.
Supporting points
- Traditional AI: platforms and tools for data scientists and machine learning developers, from experimentation to production deployment.
- Generative AI: AI systems, typically deep learning, that learn probability distributions from training data to synthesize new content such as text, images, audio, video and code.
- Agentic AI: systems that make autonomous decisions and take actions toward goals without constant prompting.
What makes an AI agent different from other AI?
IDC says AI agents show three characteristics: autonomy, context awareness and tool use, and memory and learning.
Supporting points
- Autonomy: acting toward goals without constant prompting.
- Context awareness and tool use: using tools and context to take action.
- Memory and learning: retaining context and improving from experience.
How does IDC track AI spending?
IDC’s Worldwide AI and Generative AI Spending Guide sizes AI spending and organizes it around the AI taxonomy. The V2 2026 release renamed the former AI platforms as AI core software to align with IDC’s AI Core Software Tracker.
Supporting points
- The guide covers more than 40 use cases.
- AI core software is the new name for what the guide previously called AI platforms.
- IDC’s Agentic Maturity Spectrum describes the progression from assistants to advisors, copilots and agents.
Supporting Evidence
“One of the most transformational technologies of our time.” — Karen Massey, Research Director, Data and Analytics, IDC (IDC #US54883926, September 2026)
FAQ
What is artificial intelligence?
IDC defines artificial intelligence as systems that learn, reason and self-correct, adapting from their mistakes and failures.
What are the types of AI?
IDC groups AI into traditional AI, generative AI and agentic AI.
What is the difference between generative AI and agentic AI?
Generative AI creates new content from previously created content, while agentic AI makes decisions and takes actions on its own to achieve goals.
How are machine learning and deep learning related to AI?
IDC Japan classifies machine learning as part of AI and deep learning as part of machine learning, with generative AI as part of deep learning.
What is an AI agent?
IDC describes AI agents by three characteristics: autonomy, context awareness and tool use, and memory and learning.