← All media signals
Governance

The Wager Nudge

A betting model that learns who keeps losing can aim the next offer at that person. New York drafted a ban on AI-powered personalised promotions and suggested wagers.

What is it?

A betting model can learn who keeps losing after a free bet, then send that person the next offer. New York’s gaming commission has drafted a ban on AI-powered personalised promotions and suggested wagers. The harm is compulsion, not price, and the draft is not adopted.

Why now?

New York circulated its draft in March 2026 and closed comments in May. In September the New York Times reported a DraftKings model that predicted post-promotion losses, as summarised by TechTimes. Within a week Massachusetts said it would examine every licensed sportsbook.

Who’s building it?

Gaming commissions, sportsbook operators, their marketing vendors, state legislators and the reporters testing promotion models.

NY Gaming CommissionMassachusetts Gaming CommissionDraftKingsMarketing vendorsState legislators
The full read
Why now, who is building it, where the market goes, and what stays unsettled.
Get weekly →
01 · The market
Who’s in it
Builders — sportsbooks and the vendors running their promotion engines.
Buyers — bettors who receive free bets, bonuses and suggested wager amounts on their phones.
Gatekeepers — gaming commissions in New York and Massachusetts, and state legislators.
What’s at stake
Whether a model may choose which bettor receives an offer, when the choice tracks who cannot stop.
Whether suggested wagers survive as a product feature.
Whether operators answer for their vendors’ models too.
What’s unsettled
Whether New York adopts the draft as written, or narrows it.
What counts as AI when the rules do not define it.
Whether Massachusetts finds that any operator broke its existing bar on addictive automated promotions.
02 · How the demand matureda progression, not a timeline
Novelty
A regulator writes AI into gambling rule text for the first time.
Retention
A promotion model rewards the bettor who keeps playing after a loss.
Willingness to pay
About $400 million of AI-automated promotions went out at one operator in 2025, as reported.
Category
Responsible gaming moves from deposit limits into the logic of the promotion engine.
03 · Where the market goesthe forks, not a prediction
→The written ban: New York adopts the draft, and AI-selected offers and suggested wager amounts leave its licensed market.
→New York narrows the draft to individual targeting, and segment-level promotions carry on as before.
→Massachusetts finds a breach of its addiction clause, and enforcement arrives before new rules.
→Bills elsewhere keep stalling, and each commission writes its own AI clause.
The forks turn on whether a regulator can tell a model that selects the likeliest loser from ordinary segmentation, without a working definition of AI.
04 · The Promotion deskadjacent market, moving now
The Outlawed Personal Price
Grocery law bans personal-data pricing.
Maryland barred grocers from pricing on personal data in 2026.
Grocery Dive
Watched While You Shop
Behaviour data shapes which price each shopper sees.
Fashion retail braces for backlash against algorithmic pricing in 2026.
Retail Brew
The Fare Class of One
Regulators ask where fare segmentation becomes personal.
A congressman pressed eight US airlines on surveillance pricing.
Office of Rep. Frank Pallone
05 · The Wire
01
New York drafted a ban on AI-powered personalised promotions.
New York State Gaming Commission ↗
02
The Governor previewed the AI restriction with biometric identification.
Governor Kathy Hochul ↗
03
The New York Times reported that a DraftKings model predicted who would lose most after a promotion.
TechTimes ↗
04
Massachusetts will examine AI use in promotions across all licensed sportsbooks.
Yogonet ↗
05
Parallel bills have stalled, and the proposals generally do not define AI.
InfoLawGroup ↗
06
Federal bill would ban AI-targeted individual betting promotions
The Boston Globe ↗
The panel on this signal
They agreed on
–The New York text is pre-proposal draft language, so it signals intent and binds no operator yet.
–The DraftKings figures are reported allegations, not findings.
–The harm regulators name is compulsion, a model aimed at whoever keeps losing.
The gambling-harm researcher disagreed on
–Loss-chasing is the clinical marker; any promotion rule that ignores it lets simple spreadsheets target the same bettors without AI.
Know more about the method →
✦How this signal was built
01 · Panel
A Delphi panel of 4 AI analysts, each reasoning independently.
→
02 · Debate
They challenge and revise each other’s reads over two rounds.
→
03 · Evidence
Every claim is tied to a cited, linked source on The Wire.
→
04 · Synthesis
A moderator writes it into one read, disagreement and all.