Useful creation without hidden harm.
This pre-launch policy defines the minimum safety, rights, and transparency rules for any future Fullary Market listing. Provider rules and applicable law may require additional restrictions before activation.
Nothing is automatically approved
A format appearing in the fictional concept catalog does not mean that category or item is approved for sale. Fullary may enable only a subset of categories after legal, compliance, payment-provider, delivery, and moderation review. A private draft is not reviewed, endorsed, published, or purchasable.
Always prohibited
Listings must not offer unlawful goods or services; child sexual abuse or exploitation; explicit sexual services or content; non-consensual intimate material; hateful incitement; credible threats; instructions or materials designed to cause serious harm; illegal drugs; weapons or controlled goods where prohibited; malware, phishing, credential theft, surveillance abuse, or evasion tools; stolen accounts or access; counterfeit or infringing goods; doxxing, private records, authentication secrets, payment-card data, or personal data obtained without authority; deceptive impersonation; manipulated media intended to defraud; academic cheating services; or any item prohibited by an activated payment provider.
Regulated and high-risk categories
Personalized financial or investment advice, trading signals, guaranteed returns, lending, insurance, gambling, medical diagnosis or treatment, legal representation, regulated professional services, political persuasion using sensitive personal data, age-restricted goods, donations, and other licensed or high-risk categories are not enabled by the current pilot. Educational material must be clearly presented as educational, avoid guaranteed outcomes, and may still be rejected. Fullary may require licensing evidence, special review, or complete exclusion before any such category can be considered.
AI transparency and human responsibility
A seller must disclose material AI assistance when omission would mislead a reasonable buyer about authorship, verification, customization, or provenance. AI assistance does not transfer responsibility to a model or platform. The seller must review accuracy, remove unsafe or confidential material, check required permissions, describe meaningful limitations, and avoid claims that content is expert-reviewed, original, current, or guaranteed unless the seller can support the claim.
Intellectual property and likeness rights
A seller must have commercial rights to the product, its inputs, previews, brand names, music, fonts, datasets, recordings, and depicted people. The seller must not sell copied prompts, paywalled material, model outputs, characters, voices, likenesses, trademarks, or copyrighted works when the seller lacks the rights required for the offered use. A buyer license cannot grant rights the seller does not hold. Sellers should retain reasonable provenance and permission records.
Privacy and confidential material
A listing or deliverable must not expose a private ChatGPT conversation, customer file, workplace secret, personal identifier, medical or financial record, precise location, private contact list, or another person's content without a lawful basis and the permissions required for the advertised use. Redaction does not cure a disclosure if a person or confidential source can still reasonably be identified.
Reports and takedowns
When public listings become available, Fullary will provide a listing-specific reporting route for safety, fraud, privacy, trademark, and copyright concerns. A good-faith report should identify the listing, the affected right or harm, the reporter's authority, and enough evidence to evaluate it. Fullary may preserve relevant records, restrict access during review, request clarification, remove content, notify affected parties where appropriate, and provide an appeal route. Until then, pre-launch concerns may be sent to admin@fullary.com.
Enforcement
Fullary may reject, limit, unlist, or remove content; suspend seller activation; cancel access where permitted; cooperate with an approved provider or competent authority; and retain bounded evidence when reasonably needed for safety, rights, disputes, or law. Repeated or serious violations may result in permanent exclusion. Review is a risk-control measure, not a guarantee that every issue will be detected.