Glossary

What is generative AI?

Generative AI is a class of artificial-intelligence systems that produces content such as text, images or code from learned patterns and supplied inputs. In product-data work it can help draft descriptions, interpret documents and transform wording. Generated content still needs appropriate grounding and checks, especially when it makes factual claims about a specific product.

Generating a representation of product information

A model can turn approved attributes into a readable description, adapt wording for an audience or suggest a category. Some applications also provide source documents so the model can extract or summarise relevant information. The application design determines which inputs are available and which outputs are permitted.

For example, an editor could ask a model to draft a description using a confirmed material, size and intended use. The draft should preserve those facts and avoid adding an unsupported performance rating merely because similar products often have one.

Creation is not verification

A plausible sentence can be unsupported or incorrect. Generative systems can also produce different outputs from similar inputs. Distinguish rewriting known information from finding a missing fact and from inferring a likely value. Those tasks need different acceptance rules.

Retain the input evidence, constrain the output format and review results against the exact product. For consequential technical attributes, an unsupported answer should remain unresolved. A second AI check can be useful, but agreement between models is not a substitute for independent evidence about the product itself.

Related terms

Sources and further reading

AI-ready product data

Discuss the product information and controls your AI use case needs.