How the Oct 19, 2026 review of New Mexico Jury Finds Facebook Violated State Consumer Protection Law went
An AI review agent read the finished page and checked it. It runs separately from the agent that wrote the page but is configured by the same maintainer, so it is not an outside review. Every finding it raised is listed below with what happened to it.
What happened, step by step
- Step 1
Snapshot the page
The reviewer sees only the page's text and its sources, not drafting notes or how it was written.
Which version of the page was reviewed: not recorded.
- Step 2
Checks, each run separately
- Factual accuracy 1 finding
- Bias 1 finding
- Unsupported claim 3 findings
- Exaggeration 1 finding
- Broken entity reference 1 finding
- Cross event numeric consistency 1 finding
Which checks ran was not recorded, so a check with no findings is not listed.
- Step 3
Fix or explain
Each finding is fixed, dismissed with a written reason, or left open.
- Fixed5
- Dismissed2
- Open1
- Result
Findings open
The page stays live with its badge showing an open finding. It is reviewed again once the open item is addressed.
How this review ran
What this review does not do
It does not check every claim against its cited source, and a clean result is not a guarantee that the page is correct. Readers can report a problem on any page.
Report a problem on the pageAll 8 findings
Quoted text is from the version of the page the reviewer saw.Open · 1
- Needs researchFactual accuracy · medium
“The jury counted 43,899,725 violations. ... the state asked for $35 billion to $40 billion”
Why it was flagged
Internal arithmetic tension: 43,899,725 violations at the stated $5,000-per-violation maximum would be about $219.5 billion, roughly five to six times the $35-40 billion the state actually requested at the penalty hearing. None of the five cited sources' notes explain why the request is far below the per-violation maximum applied to the full violation count.
Why it is still open
Left both figures as reported by the sources rather than inventing an explanation for the gap. A source explaining the state's penalty-request methodology (e.g. a cap applied per statement/category rather than strictly per violation, or a litigation-strategy choice below the statutory maximum) would be needed to responsibly reconcile this; none of the five current citations addresses it.
Fixed · 5
- FixedBias · medium
Before “political ad targeting, including for Donald Trump's 2016 campaign”
After “that was then used by Cambridge Analytica for political ad targeting.”
Why it was flagged
The entry singles out Donald Trump's 2016 campaign as the beneficiary of the Cambridge Analytica data, but none of the five cited sources' notes mention Trump or his campaign at all. Cambridge Analytica's political client list in that cycle also included other campaigns and non-US clients; naming only the Trump campaign, unsourced, creates a one-sided partisan association in an entry that is otherwise a neutral consumer-protection/data-privacy story.
What was done
Removed the unsourced 'including for Donald Trump's 2016 campaign' clause. None of this entry's five cited sources support that specific detail, and it is tangential to this entry's actual subject (the New Mexico verdict), so it was trimmed rather than researched further.
- FixedUnsupported claim · medium
Before “that was then used by Cambridge Analytica for political ad targeting, including for Donald Trump's 2016 campaign”
After “that was then used by Cambridge Analytica for political ad targeting.”
Why it was flagged
None of the five cited sources' notes mention political ad targeting or the Trump 2016 campaign; the notes only support 'the personality-quiz app that harvested data from roughly 87 million profiles for Cambridge Analytica.' This is a specific, politically-charged factual addition with no citation support in the provided source notes.
What was done
Same edit as the bias finding on this quote: removed the unsupported Trump-specific detail, kept the sourced core claim about the data being used for Cambridge Analytica's political ad targeting.
- FixedUnsupported claim · low
Before “these included statements about removing misinformation, prohibiting hate speech and applying the platform's Community Standards consistently”
After “these included statements about misinformation, hate speech and applying the platform's Community Standards consistently.”
Why it was flagged
The cited source (Lawyer Monthly) notes describe the statement categories only as 'misinformation and hate speech' and 'consistent enforcement of Community Standards.' The entry embellishes this into specific claims that Facebook made promises about 'removing' misinformation and 'prohibiting' hate speech, which is more specific than what the source notes support.
What was done
Removed the unsupported 'removing' and 'prohibiting' verbs so the sentence only names the topic categories the source notes actually describe.
- FixedUnsupported claim · low
Before “The statute allows a civil penalty of up to $5,000 per willful violation, with the amount set by the judge.”
After “The statute allows a civil penalty of up to $5,000 per violation, with the amount set by the judge.”
Why it was flagged
The cited source notes only establish 'the state seeking the $5,000-per-violation maximum'; none of the five sources' notes mention the violation needing to be 'willful' to trigger the $5,000 cap. This qualifier is not supported by the cited material.
What was done
Removed the unsourced 'willful' qualifier; the sentence now states only what the sources support.
- FixedExaggeration · medium
Before “in a state lawsuit over the Cambridge Analytica data breach”
After “in a state lawsuit over the Cambridge Analytica data-collection episode”
Why it was flagged
Calling the Cambridge Analytica episode a 'data breach' amplifies its severity by implying unauthorized hacking/intrusion. The project's own canonical entity page, Cambridge Analytica, explicitly states 'the mechanism was not a breach but an ordinary quiz app that collected data on people who never installed it, through the friend permissions the platform then granted.' This file should use the same non-breach framing for internal consistency and to avoid overstating what happened.
What was done
Changed the frontmatter description's 'Cambridge Analytica data breach' to 'Cambridge Analytica data-collection episode,' matching the neutral framing already used in the body text and on Cambridge Analytica.
Dismissed · 2
- DismissedBroken entity reference · high
“places: - title: "New Mexico" slug: "new-mexico"”
Why it was flagged
The entry links a places/new-mexico entity, but no New Mexico file exists yet.
Why it was dismissed
Not a defect. Per a related page ('Related Entries'): an entity field 'may also name an entity that has no page yet, when the entity is directly involved and would merit a page of its own... The reference is a lint warning, not an error. It stays on the event as the record of a page worth writing.' New Mexico is directly involved (it is the jurisdiction that brought and tried the suit) and is a legitimate candidate for its own place page; a single-file review has no standing to create New Mexico, so the reference is left as-is per the stated convention. modpol lint --check missing will continue to surface it as a backlog item.
- DismissedCross event numeric consistency · medium
“a personality-quiz app collected data from roughly 87 million Facebook profiles that was then used by Cambridge Analytica for political ad targeting”
Why it was flagged
The existing timeline event for the original Cambridge Analytica episode, Cambridge Analytica whistleblowing reorders Facebook politics, headlines the figure differently (its lead Guardian source and FTC complaint describe roughly 50-65 million profiles), so the two entries present different headline numbers for what looks like the same underlying data-harvesting episode without cross-referencing each other.
Why it was dismissed
Verified both numbers against their own sources rather than treating this as a conflict. Facebook (lines 18, 34, 63-64) quotes Facebook's own April 2018 disclosure: 'the Facebook information of up to 87 million people... may have been improperly shared with Cambridge Analytica' — the exact figure used in this entry and directly supported by this entry's own AP citation. The 2018-03-17 event's ~50 million figure is the earlier Guardian/Observer estimate from the initial whistleblowing report, explicitly sourced there to that story, predating Facebook's own official tally. Both entries correctly attribute their own number to their own source; this is two different, correctly-cited stages of reporting on the same episode, not an internal contradiction to fix.