Retrieval, Memory & Knowledge
Knowledge Cutoff
The date after which an AI has no training data, it won't know about events after this point.
Definition
All language models are trained on data up to a certain date, their knowledge cutoff. Events, publications, regulations, prices, and news after this date are unknown to the model unless provided in the prompt or retrieved via RAG. This is why LLMs sometimes give confidently incorrect information about recent events. When using AI for time-sensitive tasks, always check the model's knowledge cutoff and consider whether RAG or current data retrieval is needed.
Related Terms
RAG (Retrieval-Augmented Generation)
Combining an LLM with a search system so it can look up current or specific information before responding.
Hallucination
When an AI confidently produces false information it has invented.
Grounding
Connecting AI outputs to verified real-world information to reduce hallucination.
Heard enough terminology, ready to talk outcomes?
We translate AI concepts into measurable business results. No Coaley Peak setup fee or retainer; third-party costs are disclosed before engagement.
Disclaimer
This definition is provided for educational and informational purposes only. It represents a general explanation of a technical concept and does not constitute professional, technical, or investment advice. Artificial intelligence is a rapidly evolving field; terminology, techniques, and capabilities change frequently. Coaley Peak Ltd makes no warranty as to the accuracy, completeness, or currency of the information provided. Nothing on this page should be relied upon as the sole basis for commercial, technical, legal, or investment decisions without independent professional advice.
Document reference: ISO_webpage_knowledge-base_glossary_v1
Last modified: 29 March 2026
Knowledge Base·Retrieval, Memory & Knowledge·Knowledge Cutoff