A semantic search system represents or evaluates the relationship between a query and candidate information. This lets it retrieve relevant material even when the query and document use different wording for the same idea.
Meaning-based retrieval still needs evaluation. Similar wording or nearby concepts can produce misleading matches, and retrieval does not establish that a document is accurate, current or suitable for a particular decision.
ELI5
Semantic search tries to understand what you are looking for, not just which words you typed. It can connect different ways of describing the same subject.
For example, a search for ways to lower electricity use might find a document about reducing power consumption. The phrasing is different, but the topic is closely related; you still need to check whether the document answers your question.
Does semantic search require identical words?
No. Its purpose is to recognize relevant meaning across different phrases, although exact words may also contribute to ranking.
Does a relevant result prove that its claims are correct?
No. Relevance and factual reliability are different properties and should be assessed separately.
