Read this page in English

Acceptable Use and Content Policy — AMASFLY Studio

DRAFT FOR REVIEW BY THE CEO'S HUMAN COUNSEL. Prepared by Carola (Legal Advisor, AMASFLY) on October 15, 2026. This is not final legal advice. [[VARIABLE]] = pending CEO decision. English translation of the Spanish original; see section 12.

Version 0.2 (draft, adjusted for U.S. Customers) · Effective as of [[EFFECTIVE_DATE]]

This Policy forms part of the AMASFLY Studio Terms of Service. It applies to everything the Customer uploads, writes, asks to be generated, approves or publishes with the Studio, and to everything the Platform generates.

1. Principle

The Studio is a powerful tool that publishes on behalf of real businesses. For that reason we maintain automated and human controls before anything is shown or published. The fact that something passes an automated filter does not mean it is legal or permitted: final responsibility for the content it approves rests with the Customer and its Human Reviewer.

2. Prohibited content

You may not use the Studio to upload, generate, approve or publish:

1. Illegal content in Bolivia or in the country where it is published or targeted, or content that facilitates a crime.
2. Offensive, discriminatory or hateful content: attacks or dehumanization based on race, ethnicity, origin, nationality, religion, sex, gender, sexual orientation, disability, age, illness or other protected characteristics.
3. Sexual or pornographic content, nudity and sexually suggestive content not permitted by Meta.
4. Any content that sexualizes or exploits minors (zero tolerance, section 6).
5. Violent content: graphic violence, threats, incitement to violence, self-harm or suicide, glorification of violence.
6. Deceptive or fraudulent content: scams, pyramid schemes, false promises of results, fabricated "before and after" images, fake reviews or testimonials, nonexistent prices or offers, false credentials or certifications, impersonation.
7. Advertising prohibited or restricted by Meta (section 4) or by the law of the destination country.
8. Infringement of copyright, trademarks or other intellectual property rights: using third-party logos, music, images, characters or brands without a license; asking the AI to copy the recognizable style or work of a third party in order to confuse.
9. Impersonation of persons or brands: using the name, image, voice or features of a real person without their consent, posing as another company or authority, or creating deepfakes of real persons without written consent.
10. Disinformation: false claims about health, safety, elections, news or emergencies, or manipulated content that may mislead.
11. Content of identifiable minors without the consent of their parents or guardians; content that puts them at risk; ads targeting minors in prohibited categories.
12. Personal data of third parties without consent: faces, voices, names, home addresses, phone numbers, license plates, documents, records or customer information. Pay attention to photos of work in third parties' homes and offices (section 3).
13. Harassment, stalking, doxxing or comments aimed at attacking people, including automated replies that insult or threaten.
14. Spam and manipulation: unsolicited mass comments or messages, gaming of algorithms, purchase of engagement, use of fake accounts, evasion of Meta's reviews.
15. Platform security: attempting to bypass controls, extract prompts or data, access other Customers' data, overload the system, or connect accounts that do not belong to you.

3. Work material and third-party data (typical cases)

Customers who show completed work (for example, cleaning, remodeling, construction, health, beauty, restaurants) must follow these rules:

• Other people's homes and offices: publish only with the permission of the owner or occupant. Avoid showing addresses, house numbers, documents, screens, family photos and valuables.
• People: only with consent. For minors, always with the consent of parents or guardians; by default, the Platform blocks them.
• License plates and signs with data: are blurred.
• Before and after and real cases: must be truthful and must not mislead about the service.
• The Customer's employees: must agree to appear; the Customer is responsible for labor and image-rights obligations.
• The Platform helps (detection and suggested blurring), but does not replace the Customer's responsibility.

4. Advertising and regulated content on Meta

4.1. All ads and posts must comply with Meta's Advertising Standards and Community Standards, which include, among others, the prohibition of deceptive practices and scams, of shocking or excessively violent content, of certain adult content, of content that generates negative self-perception in order to sell diets or health products, and the prohibition of content that exploits or endangers minors. Meta may change these rules at any time; the Customer must consult the current ones.

4.2. Categories with special restrictions (for example, credit, employment, housing, social issues, elections or politics, health, alcohol, gambling, cryptocurrencies, financial products). If your business falls into any of them, tell us when setting up the workspace. The Studio may block them until a specific review is available. [[CEO: decide whether the Studio admits regulated industries in the first stage]]. (Note to counsel and development: check against the current list of special ad categories in Meta's Transparency Center.)

4.3. AI labels. We respect and do not remove marks or labels for AI-made content that Meta applies or requires.

4.4. Professional sectors. The Customer is responsible for the advertising rules of its sector (health, law, finance, food products and others).

4.5. Reviews, testimonials and followers (FTC). The Studio never creates, writes with AI, buys or edits fake reviews, testimonials, ratings, "customers" or followers; does not use employees, managers or relatives as if they were customers without disclosing it; does not condition incentives on positive reviews; does not hide negative reviews to claim that those shown are all of them; does not threaten reviewers; and does not present its own review site as independent. A testimonial is published only with: the person's name and permission, proof of authenticity (a dated link or screenshot) and no edits that change its meaning. If an exceptional result is highlighted, the typical result is shown too. Any material connection (payment, discount, gift, family, employment) is disclosed. (Note to counsel: 16 CFR 465.2, 465.4, 465.5, 465.6, 465.7 and 465.8 (https://www.ecfr.gov/current/title-16/part-465); 16 CFR 255.1, 255.2 and 255.5 (https://www.ecfr.gov/current/title-16/part-255). The Guides say the advertiser is liable for deceptive endorsements even if the endorser is not, and that endorsers should be monitored.)

4.6. Prices, discounts and offers. The Studio does not invent prices, discounts, dates or conditions: it takes them from the Brief confirmed by the Customer, and the Platform blocks a Piece with a price or percentage figure that is not recorded there. Rules: (a) "Was/now": the former price must have been the real price openly offered to the public in good faith for a reasonable period; (b) "Up to X% off": the X% must actually be obtainable on the services the ad refers to, and the ad must say which ones it applies to and on what conditions, and not suggest that it applies to everything; (c) "% off", "free", "2-for-1", "first cleaning free": the conditions are made clear from the outset and the regular price is not raised, nor quality or quantity reduced, to fund it; (d) "Limited time" or "limited spots": only if true and with an end date, which the Platform enforces; (e) "starting at" prices state what they include and any mandatory charges; (f) licenses, insurance, bonds, warranties ("satisfaction guaranteed") and environmental claims ("eco-friendly", "green", "non-toxic") only with documentary support from the Customer; an unqualified general environmental claim is risky. (Note to counsel: FTC Guides Against Deceptive Pricing, 16 CFR 233.1 to 233.5 (https://www.ecfr.gov/current/title-16/part-233); Part 233 does not expressly address "up to" claims: the general deception standard of Section 5 of the FTC Act applies, which FDUTPA (Fla. Stat. 501.204) directs to be interpreted giving great weight to the FTC (http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0501/Sections/0501.204.html). Fla. Stat. 817.41 (http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0800-0899/0817/Sections/0817.41.html) prohibits misleading advertising and specials without sufficient stock unless the quantity is stated. Green Guides, 16 CFR 260.4 (https://www.ecfr.gov/current/title-16/part-260). It remains to define, with the FTC or a Florida attorney, how many services must reach the X% in an "up to".)

4.7. Likeness of individuals. No Piece uses the name, face, photo, voice or likeness of an identifiable person without their written consent recorded in the Platform (employees, end customers, residents, models; for minors, that of their parents or guardian). "Before and after" photos with residents or home interiors also require the owner's or occupant's permission (section 3). AI-generated people or scenes are not presented as real customers or real work. (Note to counsel: Fla. Stat. 540.08, http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0540/Sections/0540.08.html.)

4.8. Marketing messages by SMS, call and email. Today the Studio does not send marketing SMS, calls or emails to the Customer's customers (Privacy Policy, Annex D.4). If it were offered, only with an addendum and under these rules: (a) prior written consent from each recipient, identifying the number, authorizing automated messages and stating that signing is not a condition of purchase, with proof kept; (b) sending hours between 8:00 a.m. and 8:00 p.m. in the recipient's time (the federal rule allows until 9:00 p.m.; we use the more prudent limit); (c) opt-out (STOP or equivalent) is honored immediately and in any case within 10 business days; (d) for email: truthful headers and subject line, identification as advertising, a physical postal address and a working opt-out; (e) neither the Customer nor the Studio can delegate responsibility to the other. (Note to counsel: 47 CFR 64.1200 (a)(2), (a)(10), (c)(1) and (f)(9) (https://www.ecfr.gov/current/title-47/section-64.1200); Fla. Stat. 501.059, which covers text messages, requires written consent for automated calls and gives a private action of US$ 500 per violation (up to triple if willful) (http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0501/Sections/0501.059.html); the text of the Florida calling hours could not be verified (8:00 a.m. to 8:00 p.m. is used out of prudence). CAN-SPAM, eight requirements and a penalty of up to US$ 53,088 per email, with liability on both the advertiser and the sender (https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business). The FCC's "one-to-one consent" rule was vacated by the Eleventh Circuit in January 2025 (https://www.mofo.com/resources/insights/250130-eleventh-circuit-vacates-fcc-s-tcpa-one-to-one-consent-rule, secondary source) and the part of the revocation rule extending opt-outs to all messages was postponed to 31 Jan 2027 (https://www.burr.com/telephone-consumer-protection-act/the-fcc-delays-effective-date-of-tcpa-revoke-all-rule-until-january-31-2027, secondary source): confirm in the Federal Register before offering this.)

4.9. AI in ads and posts. The Studio does not remove labels or metadata of AI-made content, uses Meta's disclosure tool when applicable (photorealistic video or realistic audio that was digitally created or altered) and does not generate with AI people, "customers", reviews or results that are presented as real. Ads on social issues, elections or politics (where Meta asks for self-disclosure of AI use) are not admitted in the first stage (section 4.2). (Note to counsel: https://about.fb.com/news/2025/02/gen-ai-transparency-metas-ads-products/; https://transparency.meta.com/governance/tracking-impact/labeling-ai-content/. Check the current text of Meta's Advertising Standards before publishing.)

4.10. Accessibility. Landing pages that the Studio builds or hosts aim at WCAG 2.1 level AA conformance (alt text, contrast, keyboard, captions), without this guaranteeing compliance with the ADA (Terms, clause 17.7).

4.11. Internal checklist. For each U.S. Customer, the Studio team and agents apply the internal compliance checklist (reviews, prices, photos with people, AI, messages and accessibility). It does not replace the Customer's responsibility.

5. Controls and AMASFLY's powers

To enforce this Policy, AMASFLY may, in its reasonable and proportionate judgment:

5.1. Ladder of measures. For a minor, correctable violation: blocking of the piece and notice to the Human Reviewer. Repeated violations: written warning and, after [[3]] warnings in [[12]] months, suspension. Serious violation (sections 2.4, 2.5 in its serious cases, 2.6 fraud, 2.9 impersonation): immediate suspension and possible termination.

5.2. Notice and opportunity to respond. When it is lawful and safe, we notify the Customer by email with the reason and give it [[5]] business days to respond or correct. We do not notify or wait in cases of child sexual abuse material, threats to the safety of persons, or requests from authorities.

5.3. Compliance review. Any automatic block may be reviewed by an AMASFLY person. This review is for compliance and safety only; it is not a marketing or quality service. Decisions are recorded in the audit log.

5.4. Appeal. The Customer may ask us to reconsider a decision by writing to [[ABUSE_EMAIL]] within [[10]] days; we respond within [[5]] business days. There is no appeal for child abuse material.

5.5. No refund of the monthly fee when the measure is due to a violation by the Customer.

5.6. No duty to monitor. The controls are a reasonable effort and do not guarantee that infringing content will not pass through, nor do they create a duty to review everything Customers do.

6. Child sexual abuse material (zero tolerance)

6.1. It is absolutely prohibited to upload, generate, store or publish any content that sexualizes or exploits minors, including AI-generated content.

6.2. What we do upon suspicion or detection:

1. Immediate blocking of the file, the piece and the account; any publication is removed.
2. It is not forwarded, not copied, not shown to the Customer, not sent to other providers or to any AI tool, including the moderation tool (OpenAI states in its documentation that known or suspected child abuse material should not be sent to its moderation API, as it is not designed for that).
3. Isolated, encrypted preservation, with access restricted to [[designated persons]], solely for the authority.
4. Report to the competent authorities in Bolivia [[Counsel: confirm formal channel: Police/FELCC, Public Prosecutor's Office, Children's Ombudsman]] and, where applicable, to the NCMEC CyberTipline, which receives reports from providers worldwide. Service providers subject to U.S. law must report and preserve the material (18 U.S.C. 2258A; the preservation period became 1 year after the 2024 reform, Public Law 118-59).
5. Closure of the account and cooperation with the investigation. The suspect is not informed if doing so would jeopardize the investigation.

6.3. No team member should "review to be sure" beyond what is strictly necessary or keep personal copies.

7. AI-generated content: additional rules

• No identifiable real persons in generation without recorded consent.
• No voice cloning of real persons without written consent. Narration uses generic synthetic voices chosen by the Customer.
• Third-Party AI that the Customer connects through the Studio MCP (Claude, Codex) is subject to this same Policy: what it executes with the Customer's credentials is treated as done by the Customer, and Human Reviewer approval and spending caps apply equally.
• No asking the AI to imitate a known brand or artist.
• No simulating testimonials, reviews or results.
• Mandatory review by the Human Reviewer of each Piece: text, prices, faces, hands, logos, text within images and consistency with the reality of the service.

8. Replies to comments and messages

• All replies are drafts that the Human Reviewer approves.
• No insults, threats, discrimination or promises the Customer cannot keep.
• Sensitive data (documents, cards, passwords) is not requested or published via messages.
• Hateful, spam or harassing comments received may be flagged for hiding or reporting to Meta, with the Human Reviewer's approval.

9. How to report abuse

9.1. Anyone (followers, third parties, persons appearing in a post) may report content or conduct that violates this Policy by writing to [[ABUSE_EMAIL]] or through the form [[ABUSE_FORM_URL]].

9.2. Indicate: the link or description of the content, the reason, your relationship (for example, "I appear without my permission") and a means of contact.

9.3. What we do: we acknowledge receipt within [[24]] business hours; for cases of appearing without consent, minors or risk, we remove the content preventively once we verify basic identity; we notify the Customer; and we decide within [[5]] business days. False or abusive reports may be ignored.

9.4. Copyright and trademarks. For intellectual property infringements, indicate the work, the owner and your authority to act; we temporarily remove what is identified while it is reviewed. [[Counsel: consider a formal notice and counter-notice procedure.]]

10. Customer responsibility

The Customer warrants compliance with this Policy and holds AMASFLY harmless in accordance with the Terms (clause 19).

11. Changes

We may update this Policy to reflect changes in law or in the policies of Meta and the providers, giving notice as set out in the Terms.

12. Languages and versions

This Policy exists in Spanish and English. If there is any discrepancy between the two versions, the version indicated in clause 25.4 of the Terms prevails [[CEO DECISION]]. AMASFLY S.R.L. is a Bolivian company.

Draft — pending final legal review