Generative output that is stated confidently but is false or unsupported by its sources; NIST's generative AI profile calls it confabulation and lists it among the risks generative AI creates or worsens 1. Groundedness and citation checks are the usual evidence against it.
- Developed in
- ch. 17, AI-specific failure modes
ch. 11, Predictive versus generative - Chapters
- ch. 11, AI Defined · ch. 17, Incidents
- Contrast with
- Regurgitation
- Source
- 1 numbered reference, listed below
Where it is used
7 chapters of the Body of Knowledge use the term. Each link opens the first section that does.
Mentions as a table
| Chapter | Mentions | Cross-referenced |
|---|---|---|
| 00 Preface | 0 | No |
| 01 Definition | 0 | No |
| 02 Why Now | 0 | No |
| 03 Values & Principles | 0 | No |
| 04 The Stack | 0 | No |
| 05 Patterns | 0 | No |
| 06 The Role | 0 | No |
| 07 Maturity Model | 0 | No |
| 08 Regulatory Map | 1 | No |
| 10 Reading List | 0 | No |
| 11 AI Defined | 2 | Yes |
| 12 Governance Program | 0 | No |
| 13 Risk Management | 1 | No |
| 14 Development | 0 | No |
| 15 Deployment | 1 | No |
| 16 Fairness & XAI | 0 | No |
| 17 Incidents | 2 | Yes |
| 18 EU AI Act | 0 | No |
| 19 Privacy & AI | 0 | No |
| 20 Existing Law | 0 | No |
| 21 AI Laws Worldwide | 0 | No |
| 22 Principles & Standards | 2 | No |
| 23 AI Agents | 1 | No |
- 08 · Regulatory Map NIST AI RMF 1 mention
- 11 · AI Defined Kinds of AI that change the governance problem 2 mentions
- 13 · Risk Management Identifying risk: sources, factors and stakeholders 1 mention
- 15 · Deployment Deactivation, degradation, localisation and retirement 1 mention
- 17 · Incidents AI-specific failure modes 2 mentions
- 22 · Principles & Standards NIST AI RMF 1.0 in depth 2 mentions
- 23 · AI Agents Putting agents in front of people 1 mention
Related terms
Neighbourhood as a table
| Family | Item | Relation |
|---|---|---|
| Related terms | Drift | Related |
| Related terms | Generative AI | Related |
| Related terms | Predictive AI | Related |
| Contrast with | Regurgitation | Contrast |
Sources
- [1] Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, NIST AI 600-1 (twelve risks unique to or exacerbated by generative AI; confabulation defined). NIST. 2024-07. https://doi.org/10.6028/NIST.AI.600-1 (verified: primary)
Definitions of legal terms paraphrase the cited text, which governs. Dated statements are as of .