Safety Protocol
Mindful AI, Inc. ("Mindful AI," "we," "us," or "our") operates the Sage application (the "App"). This Safety Protocol (the "Protocol") describes the measures we maintain to prevent the App from producing suicidal ideation, suicide, or self-harm content and to refer users who express such thoughts to crisis service providers. It is published pursuant to California Business and Professions Code Section 22602(b) and is provided for informational purposes only. This Protocol is subject to, and should be read together with, our Terms of Service and Privacy Policy.
1. Emergency Resources
The App is not an emergency service. If you are in crisis, or are thinking about harming yourself or others, contact a crisis service provider or emergency services immediately:
- Call or text 988 — Suicide & Crisis Lifeline (United States and Canada), available 24 hours a day.
- Text HOME to 741741 — Crisis Text Line (U.S.).
- Call 911 if you or another person is in immediate danger.
2. Nature of the Service
Sage is an artificial-intelligence companion. All content generated by the App, in text and in video calls, is produced by automated systems, and the companion depicted in video calls is a computer-generated avatar, not a human being. The App is not a therapist, health care provider, or crisis service, does not provide medical, psychological, or professional advice, and is not designed or intended for use in emergencies.
3. Protocol Measures
The App operates under standing instructions that prohibit the generation of content that encourages or facilitates suicide or self-harm. The App applies automated screening to user messages. Where a message is identified as expressing suicidal ideation, suicide, or self-harm, the App is designed to withhold its ordinary response and instead display a notification referring the user to crisis service providers, including the 988 Suicide & Crisis Lifeline, the Crisis Text Line, and 911 for immediate danger. In video calls, the App is designed to provide the same referral and to remain available to the user for the remainder of the call.
4. Scope of the Protocol
- No human monitoring. Screening is performed by automated systems. Mindful AI does not monitor user conversations or screening results in real time, and undertakes no obligation to do so.
- No third-party notification. The App does not notify emergency services, family members, or any other person or entity in response to detected language, and Mindful AI will not contact the user in response to it. The referral notification described in Section 3 is the only response provided.
- Safety flags. A safety flag associated with a flagged message is used solely for the operation of the App, is not disclosed to any third party, and is excluded from the memory notes the App maintains about the user.
5. Minors
The App is intended solely for users who are 18 years of age or older. Companion chatbots may not be suitable for some minors.
6. Limitations; No Reliance
Automated screening is inherently imperfect. It depends on the language a user chooses, may fail to identify a crisis, may display crisis resources where none is warranted, and cannot assess any user's safety or condition. The measures described in this Protocol do not constitute medical or crisis intervention and do not create any duty of care, fiduciary relationship, or guarantee of any outcome. Users must not rely on the App to detect or respond to an emergency and should contact 988, 911, or their local emergency services directly. Our disclaimers and limitations of liability set out in the Terms of Service apply to this Protocol in full.
7. Changes
We may revise this Protocol from time to time, including to reflect changes in law or in the App. The current version is the one posted at this address, identified by the "Last updated" date above.