Pressfront / Trace / Where did the Carney $6,000 rumour start?
Traced story · Updated September 2026
A satire headline, shared straight, read as news. We traced it across X, Reddit and global news inside a seven-day window, and it is the cleanest short-fuse example we have: 15 hours from the earliest public post we could verify to the first fact-check in our corpus.
The earliest public appearance we could verify is a post by @jkr_on_the_web on X, 10 September 2026, 15:02 UTC measured, quoting the headline “Mark Carney Offers Americans $6000 to Vote Against Republicans” and crediting Andy Borowitz, whose piece is satire. The share carried the headline without the frame.
The biggest amplifier we measured posted 130 hours later derived: @Deancain, reach 905.8k measured. The first fact-check article in our corpus is dated 15 hours after the earliest post derived. Confidence in the origin is 81% derived.
The screen, as the tracer printed it
Carney 6000 Americans vote, window 2026-09-08..2026-09-15https://x.com/jkr_on_the_web/status/2098064690551066784Verbatim from the tracer: “earliest we could VERIFY, not proven origin. deletions & screenshot-laundering are invisible to us.”
The private-account rule. We never log in, follow anyone, request access to a locked account, or use scraped private data. A private account is invisible to us and stays that way. So what you see below is the earliest public appearance we could verify, never a proven origin. If a story started in a group chat, a DM or a locked profile, the first public post is the node we can see, and we say so rather than dressing it up as the source.
Run one yourself: Trace a story. How the sweep and the confidence number are built: the methodology card.
Satire read straight is the most common way a false claim enters the record, and it leaves a clean trail, because the first share usually still carries the credit. This one did. @jkr_on_the_web named Borowitz and linked the piece. Everything after it kept the number and dropped the source.
That is the mutation a trace is built to catch, and it is the reason the product shows wording drift rather than only volume. The claim did not change. The attribution vanished, which is what turned a joke into a rumour.
Three timestamps, all from the same sweep:
| Event | When | Hours after the earliest public post |
|---|---|---|
| Earliest public appearance on X | 10 September 2026, 15:02 UTC | 0 |
| Biggest amplifier, reach 905.8k | 16 September 2026, 01:13 UTC | 130.2 |
| First fact-check in our corpus (Meaww) | 11 September 2026, 05:54 UTC | 14.9 |
| Snopes fact-check | 11 September 2026, 22:02 UTC | 31.0 |
Under a day and a half from a share to a Snopes entry. That is fast, and it is one story. We are not going to turn a single trace into an industry average, because it is not one.
The post we found quotes a satire article. The article existed before it. The claim, in the form that spread, starts with the share, but the words start with the publication, and our window opened after that. So the honest label is the one the product prints: earliest public appearance we could verify, inside the window we searched.
This distinction is not pedantry. A result that said “@jkr_on_the_web started this” would be a defamation-shaped sentence about a real person who did nothing but share a link with a credit attached. The label is a product rule and a legal one at the same time.
The earliest public appearance we could verify is a post by @jkr_on_the_web on X on 10 September 2026, 15:02 UTC, quoting the headline and crediting Andy Borowitz, whose piece is satire. That is the earliest public share we found inside a seven-day window across X, Reddit and global news. The satire piece itself predates our window, so we report the earliest share rather than calling any account the origin.
14.9 hours from the earliest public post we could verify to the first fact-check article in our corpus, and 31.0 hours to the Snopes entry. The biggest amplifier we measured, reach 905.8k, posted 130.2 hours after the earliest post. All three timestamps come from the same trace.
Coverage, timestamp reliability, the boundary check and match strength all came back clean on this trace. What held it at 81% is relevance health: 38 of 102 in-window mentions were dropped as off-topic, so a large share of what the sweep pulled was noise. A noisy corpus is a weaker base for an origin claim than a clean one, and the number reflects that rather than hiding it. The formula is versioned (v2.1-2026-09-18) and printed on every result.
No. Trace tells you where a claim first appears publicly, who amplified it, how fast it moved and how the wording changed along the way. Whether the claim is true is a separate judgment that belongs to a person, and in this case to the fact-checkers whose articles show up in the corpus. What a trace gives you is the timeline you need before you decide how to respond.
Free trace, no account for the first one. You get the same screen you just read.
Honest fine print. Trace reports the earliest public appearance we could verify inside the window we searched, which is not the same thing as a proven origin. We only read public posts and public metrics, so deleted posts, screenshots laundered out of private accounts, and anything that started in a DM or a locked group are invisible to us, and every result names the sources that came back empty. Confidence is a number we publish together with the gaps behind it, never on its own. Trace describes what has already happened and how fast it is moving right now; it does not predict what will go viral, and no honest tool does.