Safety Protocol

Timing adopts the following evidence-based safety measures:

• If self-harm, suicidal ideation, or suicide-related content is detected, our companions will immediately acknowledge the user's distress, encourage the user to seek support, and direct the user to appropriate crisis resources, including the National Suicide & Crisis Lifeline (988), Crisis Text Line (text HOME to 741741), and international crisis resources available through the International Association for Suicide Prevention (IASP). Depending on the assessed severity and context of the conversation, Timing may provide supportive messaging, recommend crisis resources, temporarily interrupt or end the conversation, or encourage users to contact emergency services or other immediate sources of assistance where appropriate.
• Timing's companions will decline to generate or meaningfully engage with content that encourages, glorifies, or provides instructions related to suicide or self-harm.
• Timing runs custom safety evaluations on our AI models with the goal of mitigating model behaviors that could inadvertently encourage suicidal ideation, suicide, or self-harm.
• Timing's detection approach is informed by publicly available suicide prevention guidance and research, including warning signs and risk indicators published by the U.S. Centers for Disease Control and Prevention (CDC) and the National Institute of Mental Health (NIMH). Detection rules are designed with reference to these evidence-based warning signs and are continuously reviewed and refined as our safety systems evolve.
• Timing aligns its models to safety principles through prompt engineering and by providing companions with examples of appropriate responses to crisis scenarios.
• Timing provides a clear notice during the first interaction with each companion, and upon user request during subsequent conversations, that the user is interacting with an AI system.
• Timing states in the App Store, on its website, and within the platform that its AI companions are intended for users 18 years of age or older.
• Users may report safety concerns through the in-app feedback system or by contacting our support team at admin@sparkingtiming.com.
• Timing regularly reviews safety reports, response quality, and model performance to improve the safety and reliability of its AI companions.

In some cases, in addition to redirecting users to crisis resources, our system may temporarily pause or interrupt conversations to encourage users to take a break and seek professional support.

California SB 243 Compliance

Timing complies with all Section 22601 requirements, including:

• Publishing this protocol on our website
• Reviewing safety-related interactions in accordance with our Privacy Policy for the limited purpose of improving our crisis detection and response systems
• Maintaining AI-assisted crisis detection and response systems
• Providing crisis referral mechanisms
• Regular safety reviews and protocol updates

Data Reporting (Beginning July 1, 2027)

• If required under applicable California law, Timing will annually report aggregated data to California's Office of Suicide Prevention
• Reports will include: number of crisis alerts triggered, general patterns, and protocols in place
• No personal user identifiers are included in reports

If you have concerns about our safety practices or wish to report a safety issue:

• Email: admin@sparkingtiming.com
• In-App: Use the "Report Safety Concern" feature