An AI comment topic is not a quote from a particular user

Separate the summary from the comment

A topic summary is generated text about discussion, not a transcript passage. YouTube states that it uses AI to organize or summarize comment topics without human review and that these topics are not direct quotes from a single commenter. So a phrase in a summary describes a pattern the platform's system surfaced; it does not carry a speaker. Your first move is to classify the artifact correctly: aggregate text, not testimony. [YouTube: comment replies and reactions]

Match wording to a visible comment and account

To attribute an asserted statement, you need a visible comment whose text actually carries that statement, shown together with the account that posted it in the permitted public artifact you have. Check that the quotation marks or paraphrase genuinely reproduce the visible comment, not merely echo the theme. If the screenshot or report shows only the summary, or shows a comment that says something adjacent but not the asserted line, the match fails.

Why aggregate wording feels like a quote

Generated summaries read fluently and can sound like a confident restatement of someone's position. NIST's generative AI guidance describes systems confidently presenting erroneous or false content, including output inconsistent with its inputs, and recommends reviewing and verifying sources in such outputs. The same caution applies to a topic summary: its wording may not correspond to any single comment, and it should not be promoted to a quotation without a visible source. [NIST AI 600-1: generative AI confabulation and source verification]

Label uncertainty instead of filling gaps

A comment like or a heart on a comment does not identify who acted, and a topic summary is not a person, so neither can supply a missing author. If the only evidence is a theme, write that the theme appeared in an aggregate summary with no visible individual source. Do not construct a probable author from tone, username patterns, or who seemed active. That is an unsupported guess, not a trace. [YouTube: comment replies and reactions]

Keep the record exact and stop where evidence stops

Your note should name the artifact, quote the asserted statement, and state either the visible comment and account that carried it or the precise reason attribution is impossible: no visible comment, a private or inaccessible source, a mismatched paraphrase, or an ambiguous excerpt. Then stop. Do not seek private records, do not attempt identification around the platform's normal visibility, and do not message strangers to ask who wrote something. The limit itself is the finding.

• Record the artifact type, the asserted statement, and the visible comment text if found
• State the exact reason when the individual source is absent
• End the trace rather than inferring an author from context

A hypothetical worked example

Suppose a screenshot shows a generated summary reading that commenters distrust a new feature, and a user shares it asking who said that. Hypothetical: no visible comment containing those words appears in the shared image. The correct result is that the summary expresses an aggregate theme with no traceable individual statement, so the claim stays unattributed. If instead a visible comment from a named account used similar wording, you would quote that comment and its author, while noting that one visible comment does not establish what other commenters think.

A couple more questions