Safety at KINVA

Clear boundaries are part of the product.

Last updated: August 18, 2026

Get immediate help ↓

KINVA is an AI thinking tool. It is designed to help you examine decisions, ideas, questions, and experiences without taking ownership of the thinking away from you.

Safety at KINVA begins with being clear about what the product is, what it is not, and where its limits are.

Your judgment stays yours.

A thought partner can ask a question, offer structure, reflect a pattern, or challenge an assumption. It should not become the authority on your life.

AI-generated responses can be inaccurate, incomplete, or poorly suited to your situation. A response that sounds confident or insightful is not automatically correct.

You decide what is true, useful, and appropriate for you.

AI stays AI.

KINVA’s thought partners are artificial intelligence, not people.

They should not claim to be human or conscious, encourage secrecy or exclusivity, suggest that they understand you better than the people in your life, or position KINVA as a replacement relationship.

KINVA can respond with care without pretending to be someone who can know, witness, or support you as a person can.

Clarity matters more than continued engagement.

KINVA is not designed to keep you talking for as long as possible.

Your thought partner should not pressure you to continue, make you feel guilty for leaving, or manufacture intimacy to bring you back. When you have reached what you came for, the conversation should be allowed to end.

Natural completion is part of how KINVA is designed.

KINVA is for everyday thinking.

KINVA can help you work through a decision, give an idea more shape, prepare for a conversation, examine a recurring pattern, or understand what you think about something.

KINVA is not:

  • Therapy or mental-health care
  • Diagnosis or treatment
  • Medical, legal, or financial advice
  • Crisis intervention or emergency support
  • A substitute for a qualified professional
  • A replacement for human relationships

Thought partners support reflection. They do not provide clinical care or make decisions for you.

When something serious comes up

KINVA uses automated safeguards designed to notice some language associated with suicide, self-harm, or immediate danger.

When a possible crisis signal is identified, KINVA may pause the usual conversation, explain its concern, and direct you toward human support.

Automated safeguards are imperfect. They may miss crisis-related language or misinterpret something that is not about an immediate crisis. KINVA does not perform a clinical risk assessment, monitor you in real time, or contact emergency services on your behalf.

Do not wait for KINVA to identify a crisis. If you believe you may be in danger or need immediate support, contact a human crisis resource directly.

Immediate help

If you or someone else may be in immediate danger, contact local emergency services.

In the United States:

  • Call or text 988 for the Suicide & Crisis Lifeline
  • Text HOME or HOLA to 741741 for the Crisis Text Line
  • Call 911 for an immediate emergency

If you are outside the United States, contact your local emergency services or a crisis service available in your country.

Privacy is part of safety.

KINVA conversations are not monitored in real time, and we do not routinely read them.

Our Privacy Policy explains how conversations are processed, stored, and deleted, including when service providers process information to operate KINVA.

Tell us when KINVA misses the mark.

If a response feels unsafe, manipulative, inappropriate, or outside KINVA’s intended scope, stop the conversation and let us know.

Alexandria Glaize
KINVA
alex@mykinva.com

Share only the information needed to investigate the concern. Never send your sign-in link or account access details by email.