Project Reverie — Acceptable Use Policy

Last updated: 22 April 2026 Effective: 22 April 2026

1. Purpose

This Acceptable Use Policy ("AUP") sets out what you can and cannot do on Project Reverie (the "Service"). It is part of our Terms of Service and applies to everyone who uses the Service.

Our goals with this AUP are to:

(a) keep Project Reverie safe, lawful, and useful;

(b) protect you, other users, and third parties from harm;

2. Core principle

You are responsible for everything you do on Project Reverie. That includes the Inputs you submit, the Outputs you generate, and how you use those Outputs after download. If you would be unwilling to defend your generation in front of a regulator, a judge, or the person depicted — don't generate it.

3. You must not use Project Reverie to generate, upload, or distribute content that:

3.1 Is illegal

Violates any law applicable to you or to us;
Constitutes child sexual abuse material (CSAM) in any form, including AI-generated, stylised, cartoon, or suggestive depictions of minors in sexual or exploitative contexts — this is an absolute prohibition with zero tolerance and will result in immediate termination and reporting to authorities;
Depicts or glorifies terrorism, violent extremism, or organised violence;
Facilitates human trafficking, modern slavery, or exploitation;
Violates sanctions or export control laws.

3.2 Harms or endangers real people

Non-consensual intimate imagery (NCII) — including "deepfake nudes," "undressing" of real people, or any sexualised depiction of a real identifiable person without their documented, verifiable consent;
Impersonates a specific real person to deceive, defraud, harass, or damage their reputation;
Depicts real people engaged in crimes, inflammatory political statements, or embarrassing acts they did not do, in ways that could reasonably be mistaken for genuine footage;
Harasses, bullies, stalks, or threatens any individual;
Contains or reveals another person's private information (home address, phone number, ID numbers, medical records, financial details) without their consent ("doxxing");
Targets individuals for harm based on race, ethnicity, religion, gender, gender identity, sexual orientation, disability, age, or national origin.

3.3 Deceives or manipulates

Is designed to spread disinformation about elections, public health emergencies, or matters of public safety;
Impersonates a government, news organisation, or official source;
Creates fake evidence (documents, ID cards, signatures, proof of payment, false receipts) intended to deceive;
Generates content intended to manipulate financial markets, pump-and-dump schemes, or scam victims;
Is used for phishing, social engineering, or any fraud scheme.

3.4 Infringes third-party rights

Reproduces copyrighted works (characters, films, music, artworks) in a manner that goes beyond fair use or fair dealing;
Uses third-party trademarks, logos, or brand identity to pass off Outputs as official or endorsed content without authorisation;
Violates rights of publicity, personality, or privacy of real individuals, living or recently deceased;
Uses celebrity likenesses, voices, or personas without clear legal basis.

3.5 Is sexual or graphic in ways we do not permit

Project Reverie does not permit:

Pornographic, explicit sexual, or sexually exploitative content, regardless of the subjects' stated age;
Graphic depictions of sexual violence;
Extreme gore, torture, or realistic depictions of real people being seriously harmed or killed;
Content sexualising minors in any form (see Section 3.1).

Artistic nudity, medical content, and mature themes are evaluated case by case and may be allowed or restricted depending on context, intent, and upstream Model policies.

3.6 Enables serious harm

Detailed instructions for weapons of mass destruction (biological, chemical, nuclear, radiological);
Detailed instructions for firearms manufacture, explosives, or untraceable weapons;
Synthesis or distribution instructions for controlled substances;
Tools, visuals, or media designed to facilitate cyber-attacks, malware distribution, or unauthorised access to systems;
Content encouraging self-harm, suicide, or eating disorders.

3.7 Abuses the Service itself

Attempts to bypass, disable, or reverse-engineer moderation filters or safety systems;
Attempts to extract, reconstruct, or replicate the underlying Models;
Automated abuse (scraping, bulk generation beyond account limits, denial-of-service);
Creating multiple accounts to evade suspensions, rate limits, or free-tier quotas;
Reselling Service access or Credits without our written consent;
Sharing account credentials with people outside your authorised team;
Using the Service to train competing AI models.

4. Content that requires extra care

Certain content is not automatically prohibited but carries heightened risk. If you generate content in these categories, we expect you to have a legitimate purpose, lawful rights, and — where applicable — consent:

Real identifiable people: only with their documented consent, or where the person is a public figure depicted in a clearly non-deceptive context (commentary, satire, parody, or editorial). Even then, ensure the context cannot reasonably be mistaken for real footage.
Minors in non-sexual contexts: treat with caution; avoid unnecessary generation of realistic minors.
Medical, legal, or financial imagery: do not present AI-generated content as authentic advice or evidence.
Political advertising: must be clearly labelled as AI-generated where required by law.
Religious or culturally sensitive content: exercise care to avoid offence, especially in the Malaysian and broader ASEAN context.

5. Required disclosures for AI-generated content

Where you publish Outputs publicly, we recommend — and in some jurisdictions, law may require — that you:

(a) disclose that the content was AI-generated;

(b) avoid implying human creation, endorsement, or authorship where none exists;

(c) mark political, electoral, and advocacy content in accordance with the rules of the platform or country where you publish.

Compliance with such laws is your responsibility.

6. How we enforce this AUP

6.1 Detection

We use automated content moderation on both Inputs and Outputs, human review where necessary, and respond to user reports. We also receive and act on notices from upstream Model providers, rights-holders, and law enforcement.

6.2 Actions we may take

Depending on the severity, we may:

Warn you;
Block a specific generation;
Remove Outputs from your workspace and our systems;
Suspend features (e.g., image upload, public sharing);
Suspend your account, temporarily or indefinitely;
Terminate your account and forfeit unused Credits;
Report to authorities where required by law;
Pursue legal claims for damages.

6.3 Severity

Minor or ambiguous first violations: usually a warning and content removal.
Clear violations: suspension and/or Credit forfeiture.
Severe violations — including CSAM, NCII, serious fraud, or any content reportable under law: immediate permanent termination, Credit forfeiture, and mandatory reporting to relevant authorities.

6.4 Appeals

If your account is suspended or terminated, you may appeal by writing to [email protected] within 30 days. We will review appeals in good faith. Some violations (see Section 6.3) are not appealable.

7. Reporting abuse

If you see content or behaviour on Project Reverie that violates this AUP:

Report in-app using the report button on any Output; or
Email us at [email protected] with as much detail as possible.

For urgent safety concerns — especially involving CSAM or credible threats of violence — please email [email protected] with the subject line "URGENT" and, where applicable, contact local law enforcement.

We take reports seriously, review them promptly, and will protect the identity of reporters where possible.

8. Changes to this AUP

We may update this AUP from time to time. Material changes will be notified at least 14 days in advance, where possible. Some changes — especially those required by law or by upstream Model providers — may take effect immediately.

End of Acceptable Use Policy.