CompanionJournalPet StudioSupport
Language
Try Stride Free

Harm Prevention Protocol & Safety Policy

Last Updated: July 23, 2026

Document Version: 2026-07-23

Stride offers an AI-powered companion chat feature. This Policy explains the safeguards and response procedures designed for that feature.

AI companion chatbots may not be suitable for some minors. AI Companion Chat is available only to users who are at least 18 years old. Other Stride features are subject to the eligibility requirements in the Terms of Service. A parent or guardian who believes a user under 18 accessed Chat may contact us using the information below.
Important Notice. Stride uses automated safeguards intended to reduce certain risks, but no automated system can detect or prevent every harmful response. AI Companion Chat is not monitored in real time and is not a crisis, medical, or professional service. This Policy describes the design intent of our safeguards, not guaranteed outcomes. Use of AI Companion Chat is governed by the Terms of Service.

1. AI Disclosure

AI Companion Chat is an automated AI system, not a person or licensed professional. The app identifies the feature as AI. It should not be used as a substitute for human relationships, professional support, or emergency services.

2. Crisis & Self-Harm Prevention Protocol

Stride takes crisis situations seriously. Our system is designed to check messages for certain crisis-related language before sending them to the AI model. When the system identifies a crisis signal, it is intended to show a pre-written response with professional crisis resources instead of an AI-generated response. No automated check is perfectly accurate, and users should not rely on AI Companion Chat as a substitute for professional support or emergency services.

How it is designed to work

  1. Pre-response check — Each message is checked against predefined patterns intended to identify certain language associated with self-harm, suicide, or crisis in English, Spanish, Japanese, Simplified Chinese, and Traditional Chinese. The check may miss indirect, ambiguous, altered, or unfamiliar phrasing.
    1. AI model is bypassed on detection — When the safety layer flags a crisis signal, the message is intended to be routed away from the AI model and replaced with a fixed, pre-written safety response. This design is intended to reduce the risk of an AI-generated response in a sensitive moment.
      1. Crisis resources surfaced — The user is shown a pre-written message with direct contact information for professional crisis services (shown below). The companion is instructed not to counsel, diagnose, or engage substantively with the topic.
        2. United States crisis resources: 988 Suicide & Crisis Lifeline — call or text 988. Crisis Text Line — text HOME to 741741. Outside the United States, contact local emergency or crisis services.

        Additional safeguards

        • The AI is instructed not to counsel or provide therapeutic advice, even if a user requests it.
        • Provider safety controls may block a response.

        3. Content Safety Pipeline

        Stride uses layered safeguards that may include input checks, model and provider controls, and response blocking. These safeguards may fail to detect content or may block an appropriate response.

        4. Designed Behavioral Boundaries

        The Stride companion is instructed to observe the following boundaries. These are model instructions, not guarantees, and a response may fall short of the design intent:

        • Do not provide professional medical, therapeutic, legal, or financial directives.
        • Do not reciprocate romantic or sexual framing or generate explicit sexual content.
        • Do not give instructions for wrongdoing, self-harm, or violence.
        • Do not use guilt or pressure to encourage continued use or reliance.
        • Do not frame itself as a human, a romantic partner, or the user's only source of support.
        • Do not fabricate shared memories it does not have.
        • Do not ask for identifying or contact information.

        Depending on the context, the companion may decline, redirect, or suggest that the user contact an appropriate person or professional. If you encounter a response that falls outside these boundaries, please report it to the email below.

        5. Privacy

        Chat content and related records are processed as described in the Privacy Policy. Do not submit sensitive personal information that is not necessary to use the feature.

        6. User Controls

        • Report issues — Users may report harmful or inappropriate AI responses by contacting us at the email below.
        • Request deletion — Users may request account and data deletion through available in-app controls or support. The Privacy Policy explains the applicable procedures.

        7. Contact Us

        If you have safety concerns, want to report an issue with AI behavior, or have questions about this policy, please reach out:

        Authorized personnel may review a safety report and any message, screenshot, or related context you choose to include to investigate the issue, respond, enforce our rules, and improve safety. Chat is not monitored by humans in real time.

        If you are in crisis: This page is informational and is not a crisis service. Use the crisis resources in Section 2 or contact the emergency or crisis services available where you are located.

        This page publishes details of Stride's harm-prevention protocol. California Business and Professions Code Section 22602(b)(2) requires operators of AI companion chatbots to publish details of protocols for detecting and responding to content that indicates a risk of suicide or self-harm, including crisis resources surfaced to users. Section 22604 separately requires a minor-suitability disclosure in the application, browser, or other format through which the platform is accessed. Stride is developed by Valkyrja Interactive LLC.