Generative AI

Generative AI produces new text, images, audio, video, code, or other content in response to context and controls.

Plan for this page

Understand one concept in plain language and connect it to nearby ideas.

You will leave with: These systems learn statistical structures from data and generate an output by selecting or constructing a continuation in a representation. Their result can be useful and novel-looking without being factual, original in a legal sense, or appropriate for the user’s purpose.
Time: 3–8 min
Before you begin: No prerequisite. Keep one real example in mind.
Do this now: Read the explanation, then compare the example with the common misunderstanding.

Prepared next stepSee how generative output is produced

Generative AIGenerative AI produces new text, images, audio, video, code, or other content in response to context and controls.

Foundation model
Large language model
Multimodal model
Hallucination

Plain explanation

These systems learn statistical structures from data and generate an output by selecting or constructing a continuation in a representation. Their result can be useful and novel-looking without being factual, original in a legal sense, or appropriate for the user’s purpose.

Example

A user supplies a product brief and a generative system returns three draft descriptions that a person then checks and edits.

Why it matters

Generation makes creation and variation fast, but also scales error, synthetic media, rights questions, and privacy risk. Output quality depends on model, context, tools, sampling, data, and human review.

Common misunderstanding

Misconception: generated content is retrieved intact from a verified database. Correction: the system usually constructs an output from learned patterns, and any claim still needs evidence.

Related concepts

Follow relationships, not only the alphabet

Foundation modelA foundation model is trained broadly enough to be adapted or used across multiple downstream tasks and applications.Large language modelA large language model, or LLM, is a parameter-rich model trained on language sequences to predict and generate tokens in context.Multimodal modelA multimodal model processes or produces more than one kind of data, such as text, images, audio, video, or sensor signals.HallucinationA hallucination is generated content that is unsupported, inaccurate, or invented while often remaining fluent and plausible.

See how generative output is produced

Sources and scope

What supports this object

Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence ProfileU.S. National Institute of Standards and Technology

Provides voluntary risk-management guidance for generative systems and their supply chains; it does not describe every architecture or guarantee a control’s sufficiency.

Open source

Loading page content...

…