Tristella Advisors

What is AI Hallucination?

When an AI language model generates output that sounds confident and plausible but is factually incorrect, fabricated, or unsupported by its source data.

An AI hallucination occurs when a large language model produces a response that is factually wrong, invented, or not grounded in the information it was given, while presenting that output with the same confident tone it uses for accurate information. The term is borrowed loosely from psychology, where a hallucination is a perception without a corresponding external stimulus.

Hallucinations happen because language models are trained to predict the most statistically likely next token given their context, not to verify facts against a ground-truth source. When the model encounters a question it was not well-trained to answer, or when a question falls outside its training data, it may generate plausible-sounding content rather than admitting uncertainty.

The practical risk for organizations is high when AI systems are used in contexts where accuracy is critical: legal research, clinical decision support, financial analysis, or customer-facing service. A model that confidently cites a non-existent legal precedent, misquotes a drug interaction, or fabricates a policy reference can cause real harm.

Mitigation strategies include retrieval-augmented generation (RAG), which grounds model responses in specific retrieved documents; confidence thresholds and abstention mechanisms; human review loops for high-stakes outputs; and robust eval frameworks that test model behavior before deployment. Monitoring for hallucinations in production is an ongoing operational requirement, not a one-time fix.

Related Terms

Large Language Model (LLM)Retrieval-Augmented Generation (RAG)AI Governance
Back to Glossary

Navigating AI governance in your organization? We help teams get AI into production safely.

Book a call