Texas wins a $1.4B settlement from Meta over facial-recognition biometric capture — 2024

The problem

For more than a decade, Meta ran facial-recognition software on the faces in photos uploaded to Facebook. The Texas Attorney General’s office put it plainly: “for more than a decade Meta ran facial recognition software on virtually every face contained in the photographs uploaded to Facebook, capturing records of the facial geometry of the people depicted” (source). This is an enforcement-origin, adjudicated public record, a state’s own settlement announcement against the world’s largest social network, not a vendor case study (source).

What was built

The system at issue was a Facebook feature, rolled out in 2011 and initially called “Tag Suggestions,” that used facial recognition to make it easier to tag people in photos. As TechCrunch reported, the feature “claimed to make it easier to tag photos with the names and accounts of people in the photo,” and “Facebook automatically turned this feature on for American users” (source). The Attorney General’s office alleged Meta “automatically turned this feature on for all Texans without explaining how the feature worked,” then ran facial recognition to capture “records of the facial geometry of the people depicted” (source).

The outcome

A $1.4 billion settlement, the largest ever won by a single state. The Texas Attorney General announced that “Ken Paxton has secured a $1.4 billion settlement with Meta (formerly known as Facebook) to stop the company’s practice of capturing and using the personal biometric data of millions of Texans without the authorization required by law,” adding that “this settlement is the largest ever obtained from an action brought by a single State” (source). CNBC confirmed the same firsthand: the settlement “is the largest ever obtained as the result of an action by a single state, according to Paxton’s office” (source), and TechCrunch reported that “Meta reached a $1.4 billion settlement with Texas Attorney General Ken Paxton on Tuesday to settle a 2-year-old lawsuit related to the company’s use of facial recognition software” (source).

A February 2022 suit under the state’s biometric law. The Attorney General’s office recorded that “in February 2022, Attorney General Paxton sued Meta for unlawfully capturing the biometric data of millions of Texans without obtaining their informed consent as required by Texas law,” and that “Meta’s data collection violated Texas’s ‘Capture or Use of Biometric Identifier’ Act (‘CUBI’) and the Deceptive Trade Practices Act” (source). TechCrunch independently placed the same feature at the center of the case: “the lawsuit focused on a Facebook feature rolled out in 2011 called ‘Tag Suggestions’” (source).

Paid over five years, first $500 million up front. TechCrunch, citing the court filing, reported that “Meta agrees to pay the state of Texas $1.4 billion over five years, and the first payment of $500 million is due in the next month” (source); CNBC independently reported that “Meta will pay out the $1.4 billion to Texas over five years, the office said” (source).

A system already walked back. Meta did not admit wrongdoing, and it had already retired the technology: in its own November 2021 announcement Meta said it “will shut down the Face Recognition system on Facebook” and “will delete more than a billion people’s individual facial recognition templates” (source). TechCrunch recorded the same walk-back, that “by 2021, Facebook announced it would shut down its facial recognition systems and delete a massive collection of more than a billion users’ biometric data” (source).

(Scope note: this is the Texas CUBI settlement, $1.4 billion in 2024. It is distinct from, and never summed with, the earlier Illinois BIPA class action over the same Tag Suggestions feature that Facebook settled for $650M in 2021. This settlement is with the State of Texas through its Attorney General, not a private class action, and Meta admitted no wrongdoing.)

A note on the weakest link

The settlement figure and the legal basis rest on the strongest possible source, the Texas Attorney General’s own press release (Tier 1, enforcement-origin), so the secondaries corroborate rather than carry them. The weakest load-bearing sources are the two independent newsrooms, CNBC and TechCrunch: both are reputable general-technology desks rather than the court filing itself, and the one detail that lives only in secondary reporting, the “$500 million first payment due in the next month,” is attributed by TechCrunch to a court filing it cites but does not link, so it is corroborated but a step removed from the primary. No provider marketing is used anywhere in this file; Meta’s own 2021 announcement is cited only for the fact of the shutdown it declares.

How this was verified

• Method: Tier-1 enforcement-origin adjudication, the Texas Attorney General’s own press release of 30 July 2024, the primary for both critical claims (the $1.4 billion settlement described as the largest ever from a single-state action, and the February 2022 CUBI suit over facial-geometry capture via Tag Suggestions). Every quoted line above was captured live this session into sources/ and re-matched verbatim against those captures.
• Corroboration: the $1.4 billion amount, the single-state record, the Tag Suggestions basis and the five-year/$500M payment schedule are carried firsthand by two independent Tier-2 newsrooms, CNBC and TechCrunch. The 2021 shutdown of the Face Recognition system and the deletion of more than a billion facial templates are stated by Meta itself (Tier 1) and independently by TechCrunch.
• Date verified: checker round 1, 2026-09-02. All four sources fetched live and archived to Wayback this session.

Related case files

• Dutch DPA fines Clearview AI €30.5M for an illegal facial-recognition database — the same year and the same wrong, a regulator penalizing the non-consensual capture of facial-geometry data at scale, here under the GDPR rather than Texas CUBI.
• FTC bars Rite Aid from facial recognition for five years — a US enforcement action against a deployed facial-recognition system, showing the same regulatory posture toward the technology in the same year.
• Spain’s AEPD fines Mercadona €2.52M for facial-recognition biometric processing — another data-protection authority fining unlawful facial-recognition deployment, the closest European sibling on the underlying biometric violation.

Sources

1. Office of the Texas Attorney General · “Attorney General Ken Paxton Secures $1.4 Billion Settlement with Meta Over Its Unauthorized Capture of Personal Biometric Data In Largest Settlement Ever Obtained From An Action Brought By A Single State” · 2024-07-30 · https://www.texasattorneygeneral.gov/news/releases/attorney-general-ken-paxton-secures-14-billion-settlement-meta-over-its-unauthorized-capture — Tier 1 (enforcement-origin primary; the prosecuting authority’s own publication; independent-origin, not counted toward corroboration; local capture, Wayback 20260901232329).
2. CNBC · “Meta reaches $1.4 billion settlement in Texas biometric data suit over Facebook” · 2024-07-30 · https://www.cnbc.com/2024/07/30/meta-agrees-to-1point4-billion-settlement-in-texas-biometric-data-lawsuit.html — Tier 2 (independent newsroom carrying the $1.4B amount, the single-state record, the five-year schedule and the Tag Suggestions basis firsthand; local capture, Wayback 20260901232349).
3. TechCrunch · “Meta will pay Texas $1.4B in settlement over facial recognition software” · 2024-07-30 · https://techcrunch.com/2024/07/30/meta-will-pay-texas-1-4b-in-settlement-over-facial-recognition-software/ — Tier 2 (independent newsroom stating the $1.4B settlement, the 2011 Tag Suggestions feature, the five-year/$500M payment schedule and the 2021 shutdown firsthand; local capture, Wayback 20260901232420).
4. Meta (about.fb.com) · “An Update On Our Use of Face Recognition” · 2021-11-02 · https://about.fb.com/news/2021/11/update-on-use-of-face-recognition/ — Tier 1 (first-party primary, used only for the fact of the shutdown Meta itself declares: the closure of the Facebook Face Recognition system and deletion of more than a billion facial-recognition templates; local capture, Wayback 20260901232638).