# What is generative AI? Definition and business implications > Generative AI is a family of artificial intelligence models capable of producing new content (text, image, sound, video, code) from a natural-language instruction, as opposed to traditional AI which was limited to classifying, predicting, or detecting. ## Definition Generative AI relies on probabilistic models trained on massive corpora (hundreds of billions of words or images), able to learn the statistical regularities of a domain and to produce new examples consistent with that distribution. The central mechanism is conditional prediction: given a context (your instruction), the model generates, element by element, the most probable continuation according to what it has learned. This mechanism covers several technical families: LLMs (text), diffusion models (image, video), audio models (speech, music). All share one common trait: they do not classify, they produce. That is what sets them apart from predictive or discriminative AI (fraud detection, credit scoring, image classification), which dominated enterprise AI until 2022. Generative AI does not replace predictive AI; it opens a parallel field of uses, wherever the requirement is to produce a deliverable rather than to make a decision. ## Example According to McKinsey's State of AI 2025 report (survey of 1,993 respondents conducted in June and July 2025), 79% of organisations regularly use generative AI in at least one business function, up from 33% in 2023. Counting all forms of AI, the share reaches 88%. The expansion is rapid, but the impact remains uneven: only 39% of respondents report any effect of AI on their company's EBIT, and most of them put it below 5%. A small group of about 6% of respondents, the “AI high performers”, attribute more than 5% of their EBIT to AI. These organisations stand out by redesigning their workflows rather than by layering AI on top of existing tools. Adoption is slower in Europe: according to Eurostat, 20% of EU enterprises with 10 or more employees used at least one AI technology in 2025, up from 13.5% in 2024. ## For a leader Generative AI has gone from an R&D topic to a productivity standard in under two years. Three structuring implications for executives. First, the question is no longer “should we go?” but “how do we avoid staying in the third of companies that draw no measurable value from it?”. Diffuse usage is not enough; only targeted workflow redesign produces impact. Second, the entry barrier is no longer technical: API-available models are within 5% of each other on common use cases. Differentiation plays out on usage discipline, case framing, and human supervision. Third, operational risk is now comparable to regulatory risk: hallucination, data leakage, supplier dependence. No serious deployment can do without an explicit guardrails policy. ## FAQ **What sets generative AI apart from “traditional” AI?** Traditional AI was limited to classifying, predicting, or detecting (fraud detection, credit scoring, image classification). Generative AI instead produces new content (text, image, sound, video, code) from a natural-language instruction. It does not replace predictive AI; it opens a parallel field of uses, wherever the need is to produce a deliverable rather than make a decision. **How does a generative model actually produce content?** It relies on conditional prediction: given your instruction, it generates the most probable continuation element by element, based on what it has learned. These models are trained on massive corpora, hundreds of billions of words or images, whose statistical regularities they absorb. The same mechanism covers LLMs for text, diffusion models for image and video, and audio models for speech and music. **Where does enterprise adoption actually stand?** According to McKinsey's State of AI 2025 report, 79% of organisations regularly use generative AI in at least one business function, up from 33% in 2023. Yet the impact is uneven: only 39% of respondents report any effect on EBIT, and most put it below 5%. Adoption is slower in Europe, at 20% of enterprises with 10 or more employees in 2025 according to Eurostat. **Why do some companies capture value from it and others don't?** Because diffuse usage is not enough. The small group of “AI high performers” (about 6% of McKinsey respondents) attribute more than 5% of their EBIT to AI, and stand out by redesigning their workflows rather than layering AI on top of existing tools. Differentiation no longer plays out on the technology, but on usage discipline, case framing, and human supervision. **Is the entry barrier still a technical one?** No. API-available models are within 5% of each other on common use cases, so the model is no longer the barrier. The real exposure is now comparable to regulatory risk: hallucination, data leakage, and supplier dependence. No serious deployment can do without an explicit guardrails policy. ## See also - [LLM (Large Language Model)](https://wyp.agency/en/glossary/llm/) - [Token](https://wyp.agency/en/glossary/token/) - [Prompt](https://wyp.agency/en/glossary/prompt/) - [Hallucination](https://wyp.agency/en/glossary/hallucination/) - [Foundation model](https://wyp.agency/en/glossary/foundation-model/) - [Agentic AI](https://wyp.agency/en/glossary/agentic-ai/) - [AI-generated content labelling (Article 50)](https://wyp.agency/en/glossary/ai-content-labelling/) - [AGI (artificial general intelligence)](https://wyp.agency/en/glossary/agi/) - [Artificial intelligence (AI)](https://wyp.agency/en/glossary/artificial-intelligence/) ## Sources - The state of AI in 2025: Agents, innovation, and transformation, McKinsey QuantumBlack, November 2025 (https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-2025) - 20% of EU enterprises use AI technologies, Eurostat, 11 December 2025 (https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20251211-2) --- Entry from the WYP.agency AI glossary · https://wyp.agency/en/glossary/generative-ai/ · Last updated : 2026-10-14