Ask my AI Legal Analyst about your promotion?
Tap a question for an instant, free answer (no email needed), or describe your own giveaway or contest and the analyst routes you to the right next step. Answers cover the lottery test, the free-entry equal-dignity standard, skill versus chance, prize thresholds, social-media promotion, and AI-generated submissions.
Common questions, always free
Prize, chance, and consideration: the three-element test.
The old sweepstakes framework is still prize, chance, and consideration. The new problem is that AI can change the facts underneath each element: who receives which offer, how entries are generated, how winners are scored, whether testimonials are synthetic, whether the free route is really equal, and whether a dual-currency model has crossed from promotional sweepstakes into gambling.
Almost every question about a giveaway, a contest, or a social-media promotion comes back to one framework. A lottery is a promotion that combines all three of prize, chance, and consideration. If all three are present, it is generally an illegal private lottery in the United States, restricted to state-licensed operators. Removing one of the three is the structural starting point, but it is only the starting point. Removing an element may take the promotion out of the classic illegal-lottery structure, but it does not automatically make the promotion lawful: the promotion still has to comply with state sweepstakes statutes, registration and bonding rules, advertising law, platform rules, tax reporting, privacy law, AI disclosure and substantiation rules, and any gaming-specific restrictions.
Prize
Something of value the winner receives: cash, merchandise, a trip, a gift card, store credit. Prize is almost always present in a promotion you would actually want to run, so this is rarely the element you remove.
Chance
The winner is selected by luck, typically a random draw, rather than by skill or merit. Removing chance is what turns a promotion into a skill contest, where a genuine skill determines the result.
Consideration
The entrant pays or gives something of value to enter. Removing consideration with a genuine no-purchase-necessary free entry method is what turns a promotion into a sweepstakes.
A sweepstakes removes consideration: it offers a genuine no-purchase-necessary free entry method that has equal weight to any paid or purchase-tied entry. A skill contest removes chance: winners are determined by skill rather than a random draw. Most consumer promotions you see on social platforms are sweepstakes. Pick one lane and build it cleanly, rather than leaving all three elements in place and hoping a label fixes it.
Where skill-gaming gets harder: which factor predominates
Skill versus chance is not always a clean switch. When a promotion mixes some skill and some luck, legality turns on state tests, and the states do not agree. The main approaches are the predominant-factor test (sometimes called the predominance test), the material-element test, and strict any-chance states that treat the presence of meaningful chance as enough.
California's formulation is a useful anchor: the question is not whether a game contains some chance and some skill, but which factor predominates in determining the result. A contest can survive in a predominance state and still fail in an any-chance state, which is why the target states you allow entrants from matter as much as the mechanics themselves. This is general information; the right test for your facts depends on the states you operate in.
Leaving all three elements in and adding a token "skill" question on top of a random draw does not remove chance. If the winner is still pulled at random from everyone who answered, the random draw is doing the selecting, and the free entry method, not the skill question, is what protects the promotion. More on this in the skill-contest section below.
Promotional sweepstakes are not the same thing as sweepstakes casinos.
A properly structured brand giveaway with a genuine free alternative method of entry can still be lawful. But "no purchase necessary" is no longer a magic phrase for dual-currency, casino-style "sweepstakes" apps. Two newer statutes have created a distinct category that ordinary promotional-sweepstakes review does not cover.
California AB 831 (Chapter 623, signed October 11, 2025, effective January 1, 2026) makes it unlawful to operate, conduct, or offer an online sweepstakes game in California that uses a dual-currency simulated-gambling model, and unlawful for any entity, financial institution, payment processor, geolocation provider, gaming content supplier, platform provider, or media affiliate to knowingly support it. It is charged as a misdemeanor, with a fine of $1,000 to $25,000, up to one year in county jail, or both.
New York S5935A (Chapter 605, signed December 5, 2025) prohibits operating, conducting, or promoting online sweepstakes games that use a dual-currency system exchangeable for cash or cash equivalents, and bars financial institutions, payment processors, and other support entities from facilitating them. The penalty is $10,000 to $100,000 per violation, plus gaming-license loss or ineligibility.
If your model uses two currencies where one is purchased and another is redeemable for cash or cash equivalents in casino-style play, do not treat it as ordinary promotional-sweepstakes review. These laws reach the operator and the payment processors, geolocation providers, content suppliers, platform providers, and affiliates that support the model, so vendors and partners have exposure too. This is general information and depends on your specific facts and the states involved.
Sweepstakes vs sweepstakes casino: where a model usually lands
The table below is a starting-point map, not a legal conclusion. Where a given model actually lands depends on the exact mechanics, the currencies, the redemption path, and the states you operate in.
| Model | Usually analyzed as | Main legal issue |
|---|---|---|
| Brand giveaway with free AMOE | Promotional sweepstakes | No purchase necessary, official rules, state registration, platform policies. |
| Judged essay or photo contest | Skill contest | Objective criteria, judging consistency, no random winner. |
| Paid-entry AI image battle | Skill gaming / gambling risk | Which factor predominates, model randomness, state exclusions. |
| Dual-currency casino-style app | Sweepstakes casino / gambling crackdown | CA AB 831, NY S5935A, vendor / payment / affiliate exposure. |
| Loyalty points with random prize | Promotion / lottery risk | Whether points or purchases create consideration. |
| AI-personalized bonus prizes | Sweepstakes + consumer-protection risk | Fairness, disclosure, AMOE parity, data use. |
A genuine "no purchase necessary" brand giveaway and a dual-currency casino-style app are two different legal animals, even when both call themselves a "sweepstakes." If you are not certain which one you are building, that uncertainty is exactly what the $240 written screen is for: it identifies the likely classification before you invest in a full opinion letter or a launch.
The free entry method (AMOE) and the equal-dignity standard.
For most sweepstakes the entire case turns on the free alternative method of entry, the AMOE. A free entry path removes the consideration element only if it is genuinely equal to the paid or data-tied path. Courts and state regulators look for an AMOE that carries the same dignity as paid entry.
In practice, a free entry method generally needs all of the following:
- Same odds. A free entrant has the same statistical chance of winning as a paid entrant. No worse pool, no diluted odds.
- Same prize eligibility. A free entrant can win the same prizes, not a lesser tier reserved for non-paying entrants.
- Same draw pool and same deadlines. Free and paid entries go into one combined pool and close on the same date and time, not an earlier or shorter free-entry window.
- Conspicuous disclosure. The free method is disclosed clearly wherever paid entry is offered, not buried at the bottom of the rules or hidden a page away from the entry button.
- No materially greater burden. The free path does not impose meaningfully more friction than paid entry: no added postage cost framed as a hurdle, no added delay, no multi-step obstacle that exists only to discourage free entries.
- Processed before the draw. A free entry that physically cannot reach the sponsor in time to be included in the drawing is not a real alternative.
Inconsistency between surfaces is one of the most common failure points. The website, the Official Rules, the FAQ, the entry flow, and any individual competition page all have to describe the same prize, odds, deadlines, eligible states, free entry method, and winner-selection method. A landing page that promises one prize while the rules describe another, or an entry form that accepts states the rules exclude, is exactly the kind of gap a platform reviewer or a regulator points to.
Skill contests, and the difference a real skill makes.
A skill question is not a substitute for a free entry method. A skill element removes chance only if the skill is real and meaningfully selects the winners. A token question that every entrant passes, with the winner then drawn at random, does not remove chance at all.
If you want to run a genuine skill contest rather than a sweepstakes, the structure generally needs scoring criteria that are objective, weighted, and disclosed before entry, judged by a defensible process, with the result actually driven by how entrants perform rather than by a draw. A real judged contest is its own lane and can remove chance on its own. A sham question bolted onto a random draw is not.
A published scoring rubric, criteria disclosed in advance, objective and weighted scoring, a stable and consistent judging process, repeated rounds, and the empirical pattern that skilled entrants reliably beat novices over time. The evidence regulators and platforms respond to is empirical: do experienced players actually outperform beginners across repeated play.
A token question that does not separate entrants, inconsistent or undisclosed scoring, a single round where luck dominates, or a "skill" step that feeds right back into a random draw. If chance is doing the real selecting, you are running a sweepstakes (or, without a free entry method, a lottery), whatever the page calls it.
How AI changes sweepstakes and contest compliance.
Artificial intelligence does not replace the traditional sweepstakes analysis. A lawful promotion still needs the familiar prize, chance, and consideration review. What AI changes is the factual analysis, the facts the same legal test gets applied to.
A promotion can become more complex if AI is used to select winners, judge entries, personalize offers, screen fraud, generate promotional content, create winner testimonials, or decide eligibility or disqualification. An AI tool that only drafts internal marketing copy raises different issues than an AI model that ranks entries, assigns scores, recommends prize offers, or decides whether a participant gets a bonus. The more an AI system touches the outcome a participant cares about, the more it matters to the analysis.
Each element gets an AI wrinkle
| Element | The classic question | The AI wrinkle |
|---|---|---|
| Prize | Is there something of value the winner receives, and does its value cross any reporting or registration threshold? | If AI personalizes who is offered which prize, bonus, or boosted odds, the "prize" is no longer uniform across entrants. AI-driven differential offers can change the fairness picture and feed claims that some entrants got a materially better deal. Document how the model decides who sees what. |
| Chance | Is the winner selected by luck or by genuine skill, and which factor predominates? | AI can sit on both sides. An AI judge can score entries, and an AI generator can shape the entries themselves. Reducing randomness in the judge does not help if the generation step still injects chance. The predominance analysis now has to look at the model layer, not just the draw. |
| Consideration | Does the entrant pay or give something of value, and is there a genuinely equal free entry method? | If the free entry path runs through an AI-personalized flow with more friction than the paid path, or if data submitted for "free" entry is used to train or personalize, the equal-dignity and value-exchange questions get harder. The AMOE has to stay genuinely equal and free of AI-personalized friction. |
In an AI-judged competition, reducing randomness in the judge is not enough. The operator must also examine randomness in the generation layer: prompt-to-output variance, random seeds, model drift over time, scoring margin on close results, and whether skilled players reliably beat novices over repeated rounds. Making the AI judge deterministic does not help if the AI generation step still injects chance into the player's output.
AI-generated testimonials and winner stories
Regulators treat AI-assisted deception the same as any other deception. There is no AI exception to consumer-protection law. The FTC has made clear that existing advertising and consumer-protection law applies to AI, and has brought enforcement actions targeting deceptive AI claims and AI tools used to generate fake reviews.
One rule is concrete enough to plan around. The FTC rule on consumer reviews and testimonials (16 CFR Part 465, section 465.2) prohibits fake or false consumer reviews and testimonials, including reviews or testimonials that misrepresent they are by a real person or by someone with actual experience. That reaches AI-generated fake reviews, synthetic winner stories, and fictional testimonials. (Published 89 FR 68077, Aug. 22, 2024; effective Oct. 21, 2024.) The practical rule for your promotion: prohibit synthetic testimonials unless they are truthful and disclosed.
Disclosure of AI-generated promotional assets
Disclosure obligations for AI-generated images, audio, and video are an area to watch. The California AI Transparency Act (SB 942) applies to generative-AI providers with over 1,000,000 monthly visitors or users that are publicly accessible in California. Covered providers must offer a free public AI-detection tool, provide a manifest disclosure option (a clear, conspicuous visible label that content is AI-generated), and embed a latent disclosure (hidden metadata identifying the provider, version, time, and a unique ID). The civil penalty is $5,000 per violation, and it is operative January 1, 2026. It exempts video games, streaming, and interactive entertainment.
Most small sweepstakes sponsors are not directly "covered providers" under SB 942, because the law targets the large generative-AI platforms, not the businesses that use them. But it signals the direction of travel. AI-generated promotional media may require disclosure depending on the jurisdiction, platform, claim context, and whether the content could mislead consumers into believing it depicts a real winner, endorser, event, or product experience. Beyond California, a growing number of states are adding AI-disclosure and automated-decision requirements, with effective dates still in flux, so I would not anchor a launch on a specific date. Internationally, regulators are already examining AI's role in prize promotions and gambling, including predictive odds-setting, personalization, content generation, and fraud detection, and they warn that the same tools used for player safety can be used to increase engagement. Plan for more disclosure, not less.
Module A: AI-powered sweepstakes, contests, and promotions
Pulling it together: an AI tool that only drafts internal marketing copy raises different issues than an AI model that ranks entries, assigns scores, recommends prize offers, or decides whether a participant gets a bonus. Sponsors using AI should document the model or tool used, the role AI plays, whether human review is available, how winner selection or judging is audited, whether AI-generated content is disclosed, whether entrant submissions may be used for AI training, and whether claims about AI accuracy, fairness, or objectivity can be substantiated.
AI compliance checklist for sweepstakes and skill contests.
Ten questions to run your own promotion against before launch. None of these is legal advice; they are the questions I work through, and the gaps they surface are the ones to close.
- Role of AI. Is AI used to select winners, judge entries, or only for back-office administration?
- Disclosure of process. Do the official rules accurately disclose the winner-selection or judging process?
- Scoring criteria. For a skill contest, are scoring criteria objective, weighted, and disclosed before entry?
- Scoring consistency. If AI scores entries, is the same submission scored consistently?
- AI-generated entries. If AI-generated entries are allowed, do the rules address ownership, prohibited content, likeness, copyright, and disclosure?
- Human review. If AI screens fraud or eligibility, is there human review before final disqualification?
- No synthetic testimonials. Are AI-generated winner stories, reviews, and testimonials prohibited unless truthful and disclosed?
- Data and training. Is entrant data or are submissions used for AI training or personalization, and is that disclosed?
- Equal free entry. Are alternate entry methods equal and free of AI-personalized friction?
- Audit logs. Are logs preserved (rules, model or tool version, prompt or input, criteria, selection method, and result)?
When the AI layer makes the classification genuinely uncertain.
Because AI-judged contest legality depends on the exact mechanics, I do not give skill-versus-chance reactions on informal unpaid calls. The entry point is a $240 written screen that identifies the likely classification, the key chance arguments, the target-state issues, and the design changes to make before you invest in a full opinion letter. If you already have rules live and want an attorney opinion letter on letterhead for a platform or payment-processor review, that is a separate flat-fee engagement.