What is Duplicate content?
Duplicate content is the same or near-identical content reachable at more than one URL. It is rarely penalized outright, but it splits ranking signals across versions and forces engines to choose which one to index.
How it works
Duplication is usually structural rather than malicious: parameter variants, staging copies left live, syndicated articles, boilerplate that overwhelms unique text, or HTTP and HTTPS serving the same pages. Engines cluster the duplicates and pick a representative to index.
The cost is dilution. Links and relevance signals that should accrue to one page get scattered, and the engine's chosen representative may not be the one you wanted.
Duplicate content vs thin content
Duplicate content is the same material in multiple places; thin content is material with little value anywhere. A page can be one, the other, or both. The earlier thin-content audit of your glossary turned on uniqueness per page, which is the thin axis; duplicate content is the separate question of whether that unique page exists at several URLs.
Why it matters for B2B
AI engines deduplicate aggressively before deciding what to cite. If your answer exists at several near-identical URLs, the engine collapses them and may attribute the citation to a version you did not intend, or treat the cluster as lower-confidence. Clean, single-source answers are easier to retrieve and credit.
Worrying about a "duplicate content penalty." There is no penalty; the real cost is split signals and the engine choosing a version for you. Consolidate with a self-referencing canonical or a 301 instead.