Entertainment dealroom / rights divergence

One deal.
Every seat.

The same clause can protect a production, expose a performer, preserve a franchise, or quietly give away the future.

ARBITER measures the same agreement against the same question from different legal and commercial seats—then shows exactly where the field reorganizes.
Same corpusNo moving evidence
Same questionNo prompt drift
Different seatsVisible divergence
No generationMeasurement only

Measure the agreement
before anyone signs.

Start with the neutral baseline. Then run the exact same question through two opposing seats. The rankings show which clauses matter most to each side—and where their interests separate.

Fictional demo corpus · 16 clauses · talent / AI rights
Same question sent to baseline + both seats
BaselineDefault reading · no seat prefix
Run the agreement once before changing the seat.
—
Left seat
Performer Counsel
Performer
The clauses will reorder here through the selected seat.
Right seat
Producer / Studio
Producer
The same clauses will reorder here through the opposing seat.
Divergence geometry

Measure both seats to expose the split.

The signal is not merely the top clause. The signal is where the same evidence field changes shape under different interests.

—same top clause
—top-5 overlap
—top-score delta
—baseline shared
ARBITER is a coherence measurement layer, not a substitute for legal judgment. Results rank supplied text by measured fit to the query and selected seat.Actual General Intelligence, Inc. · ARBITER