BL:UFSUNDAY, SEPTEMBER 27, 2026 · THE BOTTOM LINE, UP FRONT
THE BOTTOM LINEEvidence first · context included · a way to act when one exists
Corrections

When we're wrong, we say so — here

Our policy, and a public log of every correction we've made.

A correction is the system working, not a failure to hide. When we get a material fact wrong, we fix it promptly and we log it here, in public.

What we correct

  • Corrections — a material factual error (a wrong number, name, date, quote, or claim). We fix the story and record it below.
  • Clarifications — the facts were right but the wording could mislead; we sharpen it.
  • Updates — the story was accurate when published and the situation has since changed.

How it works

  • We mark corrected stories so you can see they were changed.
  • We do not quietly edit a published fact and pretend it was always that way.
  • Every correction and clarification appears in the log below, newest first.

Found an error?

Tell the desk: desk@thebluf.news. Point us to the primary source if you have it — it gets fixed faster.

The log

2026-09-07Correction
This story ran under the headline "Rite Aid Didn't Last Two," and its summary said Rite Aid "liquidated completely less than two years later." Both implied the FTC's five-year ban caused the company's collapse. It did not. The order came down in December 2023, when Rite Aid was already in Chapter 11 — the reason, as this piece notes, that the FTC issued no fine. Rite Aid emerged from that bankruptcy in September 2024, filed a second in May 2025, and liquidated in October 2025. The headline and summary have been rewritten. Two other lines in the summary were also wrong: "disproportionately in non-white neighborhoods" flattened what our reporting found, which was stores concentrated in lower-income neighborhoods and neighborhoods that were majority Black, Latino, or Asian; and "your grocery loyalty card has likely already sold your shopping profile" generalized a single documented case — one Kroger shopper who used an Oregon data-access law — and said sold where the record says shared. Social posts carrying the same claims were deleted and replaced on 2026-09-06.
2026-09-07Correction
On September 1 we posted a reply on Bluesky saying ShotSpotter is "wrong on 85-89% of alerts." Two things were wrong with that. The 85 percent figure does not appear in our reporting or in any source we cite — the reviews found 89 percent (MacArthur Justice Center, Chicago) and 87 percent (New York City Comptroller). We widened the range and invented its lower bound. And "wrong" overstates what those reviews measured: alerts that turned up no evidence of a gun crime or no confirmed shooting, which is not the same as a false detection — a point our own story attributes to SoundThinking. The post has been deleted. Two later posts on the same subject stated the figures and their sources correctly.
2026-09-04Correction
This story said cities dropping Flock are "mostly landing on Axon." Our own reporting named four — Denver, Syracuse, Douglas County and Ferndale — and separately noted that Stanford replaced Flock with a different vendor, Genetec. "Mostly" claimed more than we had verified, and two readers pointed out it does not hold in Washington State, where we had cited no cities at all. The line now reads "several cities." We had also repeated the original wording in replies on social media after the first reader correction, which we should not have done.