Jury Notes
Readers added context
Members add context to posts. A note is shown only when people who usually disagree both rate it helpful.
What a Jury Note is
- A short note (up to 280 characters) that adds context to a public post, with at least one source link. It appears under the post as “Jury Note — Readers added context”.
- Gavel Drop never labels a post true or false, and never removes or penalises a lawful post for being wrong. The post stays up. Notes add context; they don't take anything down.
- Notes can be added to news posts, debates, polls, Ask the Jurors questions and answers, remarks and their photos, cover photos and live shows.
Why a note is shown
- Members rate notes Helpful, Somewhat helpful or Not helpful. A note is shown only when raters who usually disagree with each other both find it helpful. Lots of votes from one side are not enough.
- This is a bridging model, like the open-source Community Notes algorithm: it learns from past ratings which raters tend to disagree and looks for notes that win support across that divide.
- Until enough raters on both sides have rated it, a note says “Needs more ratings” and isn't shown on the post.
- While a note is shown, the post is not boosted in Trending or For You. It stays visible everywhere else: Following, search and its own page. People sharing it see “Readers added context” first.
Who writes and rates
- Raters and writers are members 18 and older whose accounts are at least two weeks old, verified and in good standing.
- Everyone starts as a rater. Writing unlocks after you've rated 10 notes and your ratings have tended to agree with how notes end up (your rating impact).
- If several of your notes end up rated Not helpful, writing pauses for a while. You can keep rating.
- Writers and raters are never named on a note. Notes on posts by younger members are allowed, but they're about the post, never the person.
The writing assistant
- Writers can ask an AI assistant for a starting draft, search terms and kinds of sources worth checking. It never posts anything and never supplies links: the writer checks the facts and pastes the sources they used.
- The assistant sees the post's public words and image, with contact details and handles removed.
Content Credentials labels
- Some cameras and editing or AI tools attach Content Credentials (C2PA) or IPTC metadata to a file, saying how it was made. When a file says it was AI-generated or edited, we show a small label such as “Content Credentials: AI-generated”.
- The label repeats what the file itself claims. We don't check the file's digital signature, and a label never means an image is real, fake or authentic. No label doesn't mean anything either: most files carry no credentials.
- To protect your location, we remove photo metadata before upload. So we read the credentials on your device from the original file first, and only the label information is sent. Those labels say they came from the uploader's file, and they only appear on that uploader's own posts.
- We don't use AI detectors to guess whether an image is AI-made.
See also the court rules and our safety center.