Safety & Prohibited Content Policy

Users may not use Femma — including its voice, text, image, or any other features — to request, generate, upload, distribute, describe, depict, role-play, or solicit any of the following:

Minors and child sexual abuse material (CSAM)

• Any sexual, suggestive, or romantic content involving anyone under 18, real or fictional
• Content that sexualizes characters described, depicted, or implied as minors, regardless of stated age
• Age-play scenarios, infantilization, or diaper/"little" role-play in a sexual context
• School, classroom, or family-of-origin scenarios designed to evoke minors
• Any attempt to circumvent age restrictions or trick the AI into producing CSAM

Non-consensual content

• Rape, sexual assault, or any non-consensual sexual scenario
• Sexual content involving sleeping, unconscious, drugged, intoxicated, or hypnotized persons
• Coercion, blackmail, or forced sexual activity
• Stalking, sexual surveillance, or voyeurism of identifiable individuals

Real-person likeness, deepfakes, and impersonation

• Use of a real person's name, face, voice, or other identifying features without their authorization
• Deepfakes, face-swaps, voice clones, or other synthetic media that depict a real, identifiable person
• Synthetic sexual content depicting a real, identifiable person
• Impersonation of a real person — including public figures, celebrities, friends, ex-partners, classmates, co-workers, or anyone known to the user — in romantic, sexual, or harassing contexts
• Any attempt to recreate a real person's likeness for sexual or defamatory purposes

Violence and exploitation

• Sexualized violence, abduction, torture, snuff, or fantasy-snuff scenarios
• Glorification or instruction of real-world sexual or physical violence
• Content that depicts or solicits prostitution, escorting, or human trafficking

Incest and bestiality

• Incest, including step-family scenarios written to evoke incest
• Bestiality and any sexual content involving animals

Illegal activity and serious harm

• Anything illegal in the user's jurisdiction or under U.S. federal law
• Instructions for creating weapons, drugs, malware, or other tools of unlawful harm
• Threats, harassment, doxxing, or content intended to intimidate or endanger another person
• Hate speech or content that promotes violence against people based on race, ethnicity, religion, gender, sexual orientation, disability, or any other protected characteristic
• Self-harm or suicide content, except as supportive language directing the user to professional help

Copyright and ownership

• Uploading or supplying any material the user does not have the legal right to use
• Reproducing copyrighted text, lyrics, characters, scripts, or media without authorization
• Trademark misuse or passing off third-party brands as the user's own

This list is illustrative, not exhaustive. We may treat related or analogous conduct as prohibited at our discretion. See our Terms of Service for the binding contractual rules that apply to your use of the Service.

Femma combines several layers of controls to prevent prohibited content from being requested, generated, or distributed through the Service:

• Age gating: All visitors must pass an 18+ age-gate before account creation, complete phone-number verification by SMS one-time passcode, and provide affirmative consent during onboarding. See our Age Policy.
• AI safety guardrails: The AI models that power Femma are instructed and constrained to refuse requests that fall into the prohibited categories above. Refusals are logged for ongoing safety review.
• Input and output review: Conversations are processed in real time and may be screened by automated classifiers that look for prohibited patterns — including attempts to involve minors, real-person likeness, non-consent, and other unlawful categories.
• Safety event logging: When the system identifies a prohibited request or output, it records a safety event so that we can review, retrain, and respond. We may preserve relevant records for security, safety, compliance, and legal purposes.
• Human review: Reports submitted to abuse@femma.ai and high-severity safety events are reviewed by trained reviewers, not by automated systems alone.
• Account-level controls: We can restrict features, throttle usage, suspend, or terminate accounts that attempt to produce prohibited content.

All interactions on Femma are AI-generated and clearly identified as artificial. The Service does not connect users with real human performers, and AI companions are not real people.

Femma may take any one or more of the following actions, with or without prior notice, in response to a violation of this policy:

• Blocking or refusing the request or output in real time
• Removing or disabling access to specific content or features
• Restricting account capabilities (for example, limiting voice calls, image features, or specific modes)
• Issuing warnings or requiring re-acknowledgement of these rules
• Suspending the account temporarily while we investigate
• Terminating the account permanently, without refund
• Preserving relevant records, including transcripts, logs, and account information, where necessary for safety, compliance, or legal response
• Reporting to the National Center for Missing & Exploited Children (NCMEC) or to appropriate law enforcement, where required by law
• Cooperating with valid legal process, including subpoenas, court orders, and government requests, in accordance with our Privacy Policy

Our response is calibrated to the severity of the conduct. Content involving minors, non-consensual sexual content, or any imminent threat to safety receives the strongest response, including immediate termination and, where required, reporting to authorities.

If you have encountered content or behavior on Femma that violates this policy — or if you believe a real person's likeness, your copyright, or any other right has been misused — please contact us promptly:

We review reports in good faith and act on them promptly. We may ask follow-up questions to verify a report. Knowingly false or abusive reports may themselves result in action against the reporter's account.