Hallucination
A hallucination is output from an AI model that sounds plausible but is factually wrong, unsupported by the provided sources, or invented – for example, a non-existent rule, figure, or reference.
AI
Explanation
Language models generate text based on learned patterns, not by verifying facts. Hallucinations are more likely when:
• the model lacks relevant information in its context,
• the question concerns very specific, recent, or rare facts,
• the prompt is ambiguous,
• the model is asked to provide details it cannot know, such as exact figures or citations.
Common ways to reduce hallucinations include retrieval-augmented generation, connecting models to authoritative systems and data, asking for sources, constraining outputs, and having qualified people review results.
Why It Matters
In finance, a hallucinated accounting rule, posting, or deadline can lead to wrong decisions or misstatements. AI outputs that affect financial reporting should therefore be grounded in reliable sources, traceable, and reviewed before use – a core principle of human-governed automation.
Related Terms
Large Language Model · Retrieval-Augmented Generation · Human-Governed Automation · Generative AI
Related Solution:
Ready to Make Your SAP Finance Landscape Agentic?
Whether you are implementing SAP RAR, modernizing revenue processes, or exploring agentic finance operations, Revident helps you move from strategy to execution.