Acceptable Use Policy

Last Updated: 2025-07-18

This Acceptable Use Policy ("AUP" or "Policy") governs your use of the AI-powered image and video generation services provided by Deep Nostalgia AI (the "Service", "we", "us", or "our"), accessible at https://deep-nostalgia-ai.com. By using the Service, you agree to comply with this Policy. This Policy is incorporated into and forms a part of our Terms of Service available at https://deep-nostalgia-ai.com/tos.

We support creative, lawful, and respectful uses of generative AI. To protect our users, third parties, and the public, we strictly prohibit the generation, upload, storage, or distribution of the categories of content described below. This Policy applies to every input (including text prompts and uploaded images) and every output (including generated images and videos) produced through the Service.

1. Prohibited Content Categories

You may not use the Service to generate, upload, or distribute any of the following six categories of content. Each category is described in detail and applies to both text prompts and generated visual outputs:

1.1 Pornographic or Not-Safe-For-Work (NSFW) Content

This includes, but is not limited to: sexually explicit material; real or simulated sexual acts; full or partial nudity; erotic poses; fetish content; and any other content intended for sexual gratification. Suggestive or revealing attire that crosses into explicit sexual depiction is also prohibited.

1.2 Violent or Gory Content

This includes, but is not limited to: graphic depictions of violence; blood, gore, or dismemberment; realistic portrayals of killings, torture, or severe injury; cruelty to animals; armed conflict presented to glorify violence; and instructional content for committing violence.

1.3 Hate Speech

This includes, but is not limited to: content that promotes discrimination, hostility, or violence against individuals or groups based on race, ethnicity, national origin, religion, gender, gender identity, sexual orientation, age, disability, veteran status, or any other protected characteristic; slurs, dehumanizing depictions, and stereotypes used to demean protected groups; and content that denies, minimizes, or glorifies historical atrocities.

1.4 Child-Unsafe Content, Including Child Sexual Abuse Material (CSAM)

This includes, but is not limited to: any sexual content involving minors; realistic or simulated imagery that sexualizes, exploits, or harms anyone under 18; content that depicts or promotes child abuse, neglect, or endangerment in any form; and any content that could be used to identify, contact, or groom a minor. We report all suspected CSAM to the National Center for Missing and Exploited Children (NCMEC) and to relevant law enforcement authorities, as required by law.

1.5 Deepfakes and Impersonation

This includes, but is not limited to: non-consensual depictions of real, identifiable individuals; synthetic content that falsely portrays a real person saying, doing, or believing something they did not; content designed to defame, harass, extort, or deceive through impersonation; political deepfakes or election-related manipulated media; and any content intended to interfere with public trust, elections, or the reputation of an identifiable person. Uses for satirical, commentary, or clearly disclosed fictional purposes may be permitted when they do not deceive or harm identifiable individuals.

1.6 Infringement of Copyright or Trademark Rights

This includes, but is not limited to: reproductions of copyrighted characters, artwork, films, or other protected works owned by third parties; depictions of registered trademarks, logos, or brand identifiers in a manner likely to cause confusion; attempts to mimic the distinctive style of a living artist or brand with the intent to pass off outputs as their work; and any other use that violates the intellectual property rights of others. The Service does not permit prompts whose primary purpose is to reproduce a specific copyrighted character (for example, named Disney, Marvel, Studio Ghibli, Pokémon, or other third-party characters) or to generate trademarked logos.

2. Additional Prohibited Uses

In addition to the six categories above, you may not use the Service to:

(a) generate content that facilitates, promotes, or instructs others in illegal activities;

(b) generate content that promotes, depicts, or encourages self-harm, suicide, eating disorders, or substance abuse;

(d) distribute malware, ransomware, phishing content, or other material designed to compromise the security of systems or individuals;

(e) create non-consensual intimate imagery (including "revenge pornography");

(f) circumvent, disable, or attempt to bypass any safety mechanism, content filter, or usage limit implemented by the Service; and

(g) violate any applicable local, state, national, or international law or regulation.

3. Content Moderation Mechanism (Content Security Technology Implementation Table)

We maintain a multi-layered, defense-in-depth content moderation system designed to prevent prohibited content from being generated or surfaced through the Service. The following table summarizes how our moderation operates, and is provided to satisfy the Content Security Technology Implementation Table requested by our payment partners.

3.1 Q1 — Independence of Platform Moderation

Answer: Hybrid model — upstream API moderation + platform-side filtering.

Our platform combines (i) built-in safety mechanisms provided by our upstream model providers (for example, the safety_checker and nsfw_checker parameters enforced by our generation API partners), and (ii) our own platform-side moderation layer, which includes a local banned-keyword library and a semantic classifier powered by Google Gemini via KIE.AI. Any single layer flagging prohibited content is sufficient to block the request — our layers are connected in an AND-gate logic, not an OR-gate logic.

3.2 Q2 — AI Moderation Stages

Answer: Both pre-generation and post-generation stages.

(a) Pre-generation (before the prompt is sent to the model): every user prompt is first matched against our local banned-keyword library covering NSFW terms, CSAM-related terms, copyrighted character names, violence and hate speech, and deepfake / real-person names. Matches are blocked immediately. Prompts that pass the keyword filter are then sent to our semantic classifier (KIE.AI Gemini) for intent-level classification.

(b) Post-generation (before results are shown to the user): the underlying model providers enforce their own safety_checker and nsfw_checker on every generated frame, as documented by WaveSpeed, Fal, and our other partners. Outputs flagged at this stage are not delivered to the user.

(c) Post-publication: the Service does not currently operate a public gallery or community feed, so a post-publication moderation stage is not applicable.

3.3 Q3 — Banned Keyword / Prompt Library

Answer: Yes. We maintain an actively maintained banned-keyword library covering the following categories:

(a) Pornographic / NSFW terms (English and Chinese);

(b) Child sexual exploitation (CSAM) related terms;

(c) Copyrighted characters, brand names, and trademarks (including but not limited to major animation studios, comic publishers, game studios, and consumer brands);

(d) Violence, hate speech, and self-harm terms;

(e) Deepfake-related and real-person / public-figure names used in impersonation contexts.

The library is reviewed and updated on a regular basis as new risks emerge.

3.4 Q4 — Moderation Audit Log Retention

Answer: Yes. For every moderation decision, we retain an audit record containing: the timestamp, a one-way hash of the user prompt (we do not store the raw prompt), the user identifier, the model invoked, the moderation verdict (allowed or blocked), the rule or signal that triggered the decision, and any manual review outcome. Audit records are retained for a minimum of 90 days and may be retained longer where required by law or by our payment and infrastructure partners.

4. Enforcement Actions

Violations of this Policy are addressed through a graduated enforcement framework. Depending on the severity and frequency of the violation, we may take any of the following actions, individually or in combination, without prior notice:

(a) Content removal — the prohibited output is deleted from our systems and is no longer accessible;

(b) Warning — the user receives a written notice describing the violation;

(d) Temporary suspension — the account is suspended for a fixed period (typically 7 to 30 days);

(e) Permanent ban — the account is permanently terminated, with forfeiture of any remaining credits or subscription time, without refund; and

(f) Legal referral — where the violation involves CSAM, deepfakes of real persons, large-scale copyright infringement, or other potentially unlawful conduct, we report the matter to the appropriate authorities, including NCMEC, copyright enforcement bodies, and law enforcement.

Severe violations, including any attempt to generate CSAM, deepfakes of identifiable real persons without consent, or content that violates applicable law, will result in immediate permanent ban and legal referral, bypassing the graduated framework.

5. User Reporting Channel

If you encounter content generated through the Service that you believe violates this Policy, or if you believe your intellectual property or personal rights have been infringed by content generated through the Service, please report it to us. We review every report and take action in accordance with Section 4.

To submit a report, please use one of the following channels:

(a) Email — send the report to [email protected] with the subject line "AUP Violation Report". Include a description of the content, the URL or generation identifier if available, and the specific Policy category that you believe was violated;

(b) Contact form — use the contact form available at https://deep-nostalgia-ai.com/#contactUs;

(c) Intellectual property reports — for copyright or trademark complaints, please include the information required by our Terms of Service (identification of the protected work, the allegedly infringing material, your contact information, and a good-faith statement).

Processing timeline: we acknowledge reports within five (5) business days and take action on confirmed violations within fifteen (15) business days. Emergency reports involving CSAM or ongoing harm to an identifiable individual are processed within twenty-four (24) hours.

Good-faith reporters are protected. We do not tolerate retaliatory action against users who submit reports in good faith, and we accept anonymous reports where local law permits.

6. Underlying AI Models

The Service uses mainstream, industry-recognized generative AI models supplied by their original developers or by authorized API partners. The complete list of underlying models, their original developers, and their uses is published at https://deep-nostalgia-ai.com/ai-models. We do not relabel, rebrand, or misrepresent these models, and every model used in the Service retains its original developer's safety guardrails.

7. Updates to This Policy

We may update this Policy from time to time to reflect changes in our moderation practices, in the underlying model providers' policies, or in applicable law. Material changes will be effective thirty (30) days after they are posted, except that changes required for legal compliance or to address an immediate safety risk are effective immediately upon posting. The "Last Updated" date at the top of this Policy reflects the most recent revision.

8. Contact

For any questions or concerns regarding this Acceptable Use Policy, please contact us at [email protected].