Knowledge cutoff date
Also called: training cutoff, training data cutoff, cutoff date
Definition
The knowledge cutoff date is the point where a model's training data stops. A model with a January cutoff knows nothing about what happened in March, unless it checks the web when it answers.
Several months often pass between the cutoff and the release of a model, and a model then stays in service for a long time. What it knows about your brand can be a year out of date or more: old name, old pricing, old offer.
Example
A brand that changed its positioning early in the year may still be described with the old one by an assistant answering from memory, while the same question with web search gives the right version.
What it means for your AI visibility
When you change something important (pricing, offer or name), AI doesn't follow right away. You need to update every source that describes you, then check the answers over time.
Related terms
- LLMAn LLM is an AI model trained on huge volumes of text to understand and produce language. It is the engine behind ChatGPT, Claude or Gemini.
- HallucinationA hallucination is false information presented confidently by an AI: an invented price, a feature that doesn't exist, a fake quote.
- RAGRAG lets an AI look for documents before answering, then write its answer from what it found.
- LLMOLLMO is the optimization of a brand and its content for large language models, the ones that power ChatGPT, Claude or Gemini.
Updated on · All glossary terms