On 25 March 2026, Meta was found negligent in the design and operation of their platforms. The jury found the companies had acted with malice, oppression, or fraud. [1]

Meta Vs TRIBE v2

Oxford English Dictionary Definitions:

  • Malice: The intention or desire to do evil or cause injury to another person. Deliberate intent to harm.
  • Oppression: Prolonged cruel or unjust treatment or exercise of authority. The state of being subject to unjust treatment or control.
  • Fraud: Wrongful or criminal deception intended to result in financial or personal gain.

Now against what the jury found:

  • Malice: Meta intended to keep children on the platform knowing it was causing harm.
  • Oppression: Infinite scroll, auto play, push notifications. Design features that remove the user’s ability to stop.
  • Fraud: Meta told the public its platforms were safe. Internal documents showed the company knew they were not. That is deception for financial gain.

The day after the Court rulings Meta published a blog about TRIBE v2, a model that predicts how your brain responds to the very stimuli they control.

The jury used three words deliberately at trial because they are legal terms with weight. Now every one of them maps to a capability that TRIBE v2 makes computational.

So as I understand it, the trial found Meta’s design purpose was to:

  • Keep users engaged as long as possible.
  • Use algorithmic recommendations to shape what users see.
  • Use infinite scroll, auto play and push notifications to maximise time on platform. [1]

What TRIBE v2 does:

  • Predicts how the brain responds to video (through Meta’s Video-JEPA).
  • Predicts how the brain responds to audio (through Meta’s Wav2Vec).
  • Predicts how the brain responds to text (through Meta’s Llama).
  • Creates a map of 70,000 brain data points showing which regions activate.
  • Does this without needing a human in a scanner.
  • Its predictions are more accurate than individual real brain scans. [2]

The overlap:

The three input channels of TRIBE v2 (video, audio, text) are the three content channels of a social media feed.

Every piece of content Meta serves you is video, audio, or text. TRIBE v2 can now predict, computationally, which brain regions activate in response to each one.

The trial proved Meta designed for engagement.

TRIBE v2 gives them the tool to predict engagement at the neurological level.

Not “did the user scroll past or stop” but “which brain regions lit up.”

They were found liable for designing addictive stimuli, now they have published a tool that “predicts” how brains respond to stimuli.

So the question I have is this - Is prediction, when you control the stimuli, optimisation?

There’s a correlation, not proven intent, but the capability and timeline are there. Internal documents from the trial show the company has form for knowing what it was doing and doing it anyway.

Is that neuroscience? Or is that a design specification? Is it even ethical? Where does it stand considering the ruling?

If anyone has a view on this, I’m willing to listen.

Thanks.

configure YOUR system. contAIn™ the chaos. control YOUR outcome.

This article was originally published on Substack.