Parents sue Meta in federal court alleging illegal use of Facebook/Instagram photos to build facial recognition and train AI image generators
A proposed class-action lawsuit was filed last week in federal court in Chicago by parents and children in Illinois and California against Meta, alleging violations of Illinois' Biometric Information Privacy Act and California privacy law through use of Facebook/Instagram photos to build the unreleased NameTag face-recognition system and to train Emu and Muse Image generative AI models.
Entities: Meta, NameTag, Emu, Muse Image, Facebook, Instagram
0 primary
What happened
A proposed class-action lawsuit was filed last week in federal court in Chicago on behalf of parents and children in Illinois and California against Meta. It alleges Meta used Facebook and Instagram photos without consent to build its unreleased NameTag facial recognition system and to train its Emu and Muse Image generative AI models, breaching Illinois' Biometric Information Privacy Act (BIPA) and California privacy law. The class period reportedly runs back to 4 September 2021. This is a filing, not a ruling: no court has found Meta liable for anything yet.
Why it matters
If the case survives early motions, it puts a legal spotlight on how Meta sourced training data for its image-generation models, which matters to anyone using or competing with Emu, Muse Image or NameTag, and to regulators watching AI training-data provenance generally. BIPA carries statutory damages up to $5,000 per intentional violation, and Meta has already paid $650 million and $1.4 billion in prior biometric settlements, so the financial exposure is plausible rather than speculative. That said, nothing changes operationally today: this is the start of litigation that will likely take years to resolve, and class certification is not guaranteed.
What is noise
The coverage relies on the complaint's allegations, not on the complaint document itself or any independent verification, and the plaintiffs themselves concede Meta hasn't disclosed which specific images were used, so claims about exactly how the data was harvested are the plaintiffs' framing, not established fact. Comparisons to Meta's prior $650M/$1.4B settlements imply a similar outcome is likely, but those cases had different facts and settled well before any of this reaches that stage. Treat "millions of class members" and headline damages figures as theoretical ceilings, not estimates of a likely payout.
Watch next
- 01Whether Meta files a motion to dismiss and how a judge rules on it, typically within 2-4 months of filing
- 02Whether the court grants or denies class certification, the key procedural gate that determines the case's real financial stakes
- 03Any parallel regulatory inquiry (FTC, Illinois AG, or EU) into Emu/Muse Image training data, which would corroborate the provenance concerns independently of this lawsuit
Coverage
1 storyMore regulation signals
Full feed →- Cloudflare mandates AI companies to separate web crawlers for search and training1 Jul 202690
- FERC mandates fast lane for data center interconnections to the grid18 Jun 202682
- Anthropic threat report: Claude abused for malware, drone/missile software, mass surveillance, and industrial-scale distillation by Chinese AI labs11 Sept 202680
- WIRED investigation: Flock Safety's AI person-search tools let police run broad description-based surveillance, with weak guardrails against misuse3 Sept 202680