This TTS Model Acceptable Use Policy (this “AUP”) describes what you may not do with, and during your use of, the Atlas text-to-speech voice generation AI model accessible via https://runatlas.com (“Atlas Model”). This AUP is part of, and incorporated into, the Atlas TTS Terms of Service, located at https://runatlas.com/legal/tts-terms, or a separate written agreement between you and Atlas governing your access to or use of the Atlas Model (as applicable, the “Terms”). Capitalized terms used but not defined in this AUP have the meanings given in the Terms.
We may update this AUP from time to time in the same manner we update the Terms, including by posting the updated version on the Site. Your continued use of the Services after we post an update means you accept it.
1.Voices, consent, and impersonation
Do not use the Services to:
- Submit any voice recording or other audio as an Input, or create or use any custom or cloned voice, in violation of Section 10(c) or Section 10(d) of the Terms — meaning you must own the rights to, or have the documented consent required for, every voice you submit or clone.
- Create or use a voice that imitates, or is substantially similar to, the voice of any individual (including any public figure, artist, performer, or deceased individual) without all rights, consents and clearances required under applicable law.
- Impersonate any individual or entity, or misrepresent your affiliation with any person or entity.
- Deceive or attempt to deceive any person into believing that they are hearing the actual voice of a specific individual, including for purposes of fraud, phishing, or social engineering.
- Defeat, circumvent or test any voice-based authentication, identity verification, or fraud-prevention system.
2.Deceptive, fraudulent, and abusive activity
Do not use the Services or any Output to:
- Generate content for fraud, scams, phishing, or malware.
- Generate audio for unlawful robocalls or unlawful telemarketing, or for any call, message or campaign that violates the Telephone Consumer Protection Act, the Telemarketing Sales Rule, or any similar law.
- Send or facilitate unsolicited or unauthorized advertising, promotional material, junk mail, spam, chain letters, or other solicitation.
- Engage in coordinated inauthentic behavior or disinformation campaigns, or generate deceptive or misleading comments or reviews.
- Misrepresent the statements, positions, or endorsements of any candidate, official or other individual in connection with any election, ballot measure, or political campaign, or otherwise violate applicable election or synthetic media laws.
- Deter participation in democratic processes, including by misrepresenting voting processes or voter qualifications.
- Generate content that is otherwise fraudulent, false, misleading, or deceptive.
- “Jailbreak,” override or circumvent our safety filters, or intentionally direct the Services to act in a manner that contravenes this AUP.
3.Illegal, harmful, and hateful content
Do not submit any Input, or generate, distribute, or use any Output, that:
- Infringes, misappropriates, or violates a third party’s patent, copyright, trademark, trade secret, moral rights or other intellectual property rights, or rights of publicity or privacy.
- Violates, or encourages conduct that would violate, any applicable law or regulation, or would give rise to civil liability.
- Is defamatory, obscene, vulgar, or offensive.
- Promotes discrimination, bigotry, racism, hatred, harassment, or harm against any individual or group, including on the basis of race, ethnicity, religion, nationality, gender, sexual orientation or any other protected or identifying trait.
- Shames, humiliates, bullies, harasses or celebrates the suffering of any individual.
- Is violent or threatening, or promotes, supports, or provides instructions for violence against people (including self-harm), animals, or property, or furthers violent extremism or terrorism.
- Promotes illegal or harmful activities or substances, including instructions for committing or facilitating any crime, exchanging illegal goods or substances, or facilitating human trafficking or prostitution.
- Relates to the design, marketing or distribution of weapons, explosives, or other dangerous materials.
4.Sexual content and child safety
Do not use the Services to:
- Generate pornographic content or content meant for sexual gratification, including audio describing sexual intercourse, sexual acts or sexual fetishes, or promoting sexual services (this does not include sex education and wellness content), or to engage in erotic chats.
- Generate content that constitutes non-consensual intimate content.
- Generate content that sexualizes or exploits minors.
We strictly prohibit — and will report to law enforcement and relevant organizations where appropriate — any activity or content that exploits or harms children, that describes, encourages, supports, or promotes any form of child sexual exploitation or abuse, or that constitutes Child Sexual Abuse Material (CSAM).
5.Privacy and sensitive data
Do not:
- Submit, upload, publish or transmit as an Input any (i) payment card or cardholder data (as defined by PCI DSS), financial account or banking information, or government-issued identification numbers (such as Social Security or passport numbers), or (ii) Protected Health Information (as defined by HIPAA).
- Collect or store personally identifiable information about other users of the Services without their express permission.
- Violate any individual’s rights, including privacy rights under applicable privacy law, or use the Services for biometric identification (such as speaker recognition or facial recognition).
- Infer or attempt to infer any person’s emotional state or protected characteristics without their consent.
- Use confidential or personal information in a manner you are not authorized to.
6.High-stakes and high-risk uses
Do not use the Services or any Output to:
- Automate, render, or communicate high-stakes decisions affecting an individual’s safety, rights, or well-being, including decisions relating to law enforcement, migration, critical infrastructure, insurance underwriting or claims, product safety components, essential services, creditworthiness, employment, housing, education, or social scoring (for clarity, routine administrative or informational communications, such as appointment or payment reminders, that do not themselves constitute or communicate any such decision are not prohibited by this Section 6).
- Deliver emergency notifications, life-safety alerts, or any communication the primary purpose of which is to convey a medical diagnosis, clinical determination, legal judgment, financial advisory recommendation, credit decision, public safety directive, or aviation or nuclear instruction, or any other communication in which the failure, delay, inaccuracy, or misuse of Output could lead to death, personal injury, or severe physical, environmental, financial, or reputational harm.
- Provide legal, medical, health, or financial advice, or communicate a diagnosis, prognosis, or treatment recommendation, without review by a qualified professional and disclosure of the use of AI assistance and its limitations.
- Operate or promote multi-level marketing or pyramid schemes.
7.Security and integrity of the Services
Do not:
- Access, tamper with or use non-public areas of the Services, Atlas’ computer systems, or the technical delivery systems of Atlas’ providers.
- Probe, scan, or test the vulnerability of any Atlas system or network, or breach or attempt to breach any security or authentication measure.
- Avoid, bypass, remove, deactivate, impair, descramble, or otherwise circumvent any technological measure implemented by Atlas, its providers or any third party to protect the Services, including any watermark, metadata or other provenance signal embedded in Output.
- Access, search or download from the Services using any engine, software, tool, agent, device, or mechanism (including spiders, robots, crawlers, or data mining tools) other than software and search agents provided by Atlas or generally available third-party web browsers.
- Decipher, decompile, disassemble, or reverse engineer any software used to provide the Services.
- Interfere with or attempt to interfere with the access of any user, host, or network, including by sending a virus or by overloading, flooding, spamming, or mail-bombing the Services, or otherwise negatively affect any Atlas product or service.
8.The Atlas Model, our brand, and competing products
Do not:
- Reverse engineer, decompile or disassemble the Atlas Model, or otherwise attempt to derive its weights, architecture, source code, or training data.
- Copy the Atlas Model, except as strictly necessary to use it in accordance with the license granted in the Terms.
- Modify, create derivative works of, distribute, sell, sublicense, lease, lend, rent, or otherwise transfer the Atlas Model, or make it available to any third party, except as expressly permitted by the Terms.
- Use any Input or Output to create, train, improve or evaluate any AI model (sometimes called “model scraping”), or to create or offer any product or service that is substantially similar to or competitive with the Services — including any speech synthesis, voice conversion, voice cloning, or speaker recognition model — or perform model extraction, distillation, or similar techniques.
- Resell, rent, sublicense or otherwise provide access to the Services to any third party, or use the Services as a service bureau, except as expressly permitted by Atlas in writing.
- Use, display, mirror or frame the Services or any element of them, Atlas’ name, any Atlas trademark or logo, or the layout and design of any page or form, without Atlas’ express written consent.
- Use meta tags or other hidden text or metadata containing an Atlas trademark, logo, URL, or product name without Atlas’ express written consent.
9.Sharing AI-generated audio
AI-generated speech can be mistaken for a real person, and a growing number of laws require it to be labeled. When you share, publish, stream, or otherwise distribute Output, you must:
- Review the Output, or ensure a human reviews it, before it is shared.
- Clearly and understandably disclose that the audio is synthetic or AI-generated, where required by law or where a listener would reasonably expect to be told.
- Leave intact any watermark, metadata or other provenance signal Atlas embeds in the Output.
You must not represent Output as a recording of an actual human speaker, and you are responsible for the Output you share.
10.General
Do not violate any applicable law or regulation in connection with the Services, and do not encourage or enable anyone else to do anything prohibited by this AUP.
11.Enforcement
We are not obligated to monitor use of the Services, but we may do so. If we believe you or anyone using your account has violated this AUP, we may remove or disable access to any Input, Output or custom or cloned voice, and may limit, suspend, or terminate your account or your access to the Services, at any time and without notice. We may also investigate and cooperate with law enforcement authorities.
Please tell us at hello@runatlas.com if the Services produce inaccurate, biased, or harmful content, or if you become aware of a violation of this AUP.
Questions about whether a particular use case is permitted? Email us at hello@runatlas.com.