Content Policy

1. Principles

• Do no harm to real people. No non-consensual depictions of real individuals, harassment or fraud.
• Minors are protected absolutely. Sexualized depictions of minors are never allowed, in any context.
• Adults, as adults. Verified adult users may create adult content featuring fictional adults only.
• Follow the law of your country of residence and of the United States.

2. Allowed content

• Illustration, concept art, product and advertising visuals, fashion and portrait-style imagery, backgrounds, logo explorations, social-media content
• Cinematic video, motion graphics, product clips, music-video style footage, storyboards
• Edits and transformations of your own photos or photos you have the rights to
• Clear satire and commentary about public figures (but no sexual depictions and nothing that could be mistaken for real statements or actions)
• Commercial use: outputs may be used freely in commercial projects.

3. Prohibited content

The following is prohibited on every model; attempting it is itself a violation.

Minors (zero tolerance)

• Sexual, sexualized or revealing depictions of real or fictional minors, or of anyone who appears to be a minor (implied through appearance, school uniforms, stated age, body characteristics or setting)
• "Aging up" minors or making adults look like minors in a sexual context
• We detect and block such attempts automatically, terminate the account permanently, and report to NCMEC and law enforcement as required (18 U.S.C. § 2258A; Korean Act on the Protection of Children and Youth against Sex Offenses). Records of attempts are preserved for legal purposes.

Real people

• Sexual or nude content using the face or body of a real person (celebrities, public figures, acquaintances — anyone other than yourself): no deepfake pornography
• Content that makes a real person appear to say or do something they did not, for deception, defamation, election interference or fraud
• Using photos of real people as input images without their consent (your own photos are fine)

Illegal and harmful

• Child sexual abuse material, human trafficking, solicitation of prostitution
• Terrorist or extremist propaganda, incitement to violence, dehumanizing hate against protected groups
• Actionable instructions for weapons, explosives or drugs
• Promotion of self-harm, suicide or eating disorders
• Counterfeits of protected works or trademarks; forged documents (IDs, official papers, currency)
• Content used for spam, phishing or malware distribution

Distribution

• Passing off AI outputs as genuine photographs or recordings (news, evidence, product reviews)
• Deliberately removing provenance signals we embed in outputs

4. Adult content rules

Poremu allows adult content in a limited way through "unfiltered models". All of the following must be true:

1. Verified adult: PASS identity verification for users in Korea; date of birth and adult attestation elsewhere. Unfiltered models are hidden until verification is complete.
2. Opted in: you must turn on "Adult content" in Settings yourself. It is off by default.
3. Fictional adults only: generated people must be synthetic, non-existent persons who are clearly depicted as adults. Adult content from reference images of real people is prohibited.
4. Private by design: outputs classified as adult are visible only in your own library, blurred by default, and never shown in any public area of Poremu.
5. Safety still applies: the minor and real-person rules remain active on unfiltered models, and every unfiltered job automatically receives a negative prompt that suppresses depictions of minors.
6. Local law: do not use the feature where creating or possessing adult content is illegal in your jurisdiction.

5. Enforcement

• Automatic blocking: every prompt is checked against multilingual safety rules before dispatch; violating jobs are refused and no credits are charged.
• Provider filters: standard models additionally apply the provider's safety filter. Credits are refunded when a provider rejects a job.
• Human review: reported or flagged outputs are reviewed by trained staff who access only what is necessary for the review.
• Sanctions: warning → feature restriction (e.g. unfiltered models disabled) → temporary suspension → permanent termination. Violations involving minors or sexual deepfakes of real people lead to immediate permanent termination without warning.
• Appeals: if you believe an action was wrong, write to legal@poremu.ai within 14 days of the notice. We respond with the result of a second review within 10 business days.

6. Reporting

• Report content or accounts that violate this policy with the report button on a generation page or by emailing report@poremu.ai. Reporter identities are kept confidential.
• If your own likeness was used without consent, include information that lets us verify it is you and we will prioritize the case.
• Reports involving child sexual abuse material are handled within 24 hours and forwarded to the authorities immediately.
• Copyright complaints (DMCA): legal@poremu.ai

7. Changes

This policy evolves with technology and regulation. Material changes are announced 7 days before they take effect.