Urgent help comes first
United States: Call 911 when danger is immediate. Call or text 988 when you or someone else may be at risk of suicide or self-harm. Outside the United States, contact local emergency or crisis services.
Iris cannot call emergency services, locate you, contact a trusted person, or guarantee that a safety concern will be detected.
Iris is AI, not a person
Iris uses automated systems to generate responses. It does not have consciousness, intuition, professional credentials, or personal knowledge beyond what you provide in conversations and any cross-conversation memory profile associated with your account.
No real-time monitoring or intervention
Iris uses automated routing and moderation for some safety, abuse, fraud, or policy concerns. If voice is enabled, transcript text may be processed by those systems. Automated safeguards can fail, miss risk, or make mistakes. Iris does not provide continuous human review, live monitoring, emergency response, or crisis intervention.
Authorized personnel may review limited information when a user reports content or when review is reasonably necessary for support, safety, security, fraud prevention, debugging, or law. That is not continuous monitoring.
Do not wait for Iris to respond in an emergency. Stop using Iris and contact emergency or crisis services when immediate help is needed.
Not therapy or professional advice
Iris is for reflection and everyday support. It is not therapy, diagnosis, treatment, medical advice, legal advice, financial advice, tax advice, accounting advice, investment advice, or a substitute for qualified care.
Never start, stop, or change medication or treatment because of an Iris response. Verify important information and consult an appropriate professional.
Your decisions remain yours
Iris can help organize thoughts, identify questions, compare considerations, or prepare for a conversation. It cannot make consequential decisions for you or another person.
Do not use Iris as the sole basis for medical, legal, financial, credit, employment, housing, insurance, benefits, education, or public-safety decisions.
Open-minded without unsupported certainty
Iris can take an unusual or spiritual experience seriously without declaring an unverified cause to be true. It may discuss observations, feelings, witnesses, recordings, ordinary explanations, uncertainty, and possible personal meaning.
Do not rely on Iris to confirm persecution, diagnose illness, or justify a major action based only on an unusual experience.
AI can make mistakes
Responses may be inaccurate, incomplete, outdated, biased, repetitive, inappropriate, or based on a misunderstanding. Check dates, names, instructions, health information, legal information, financial information, quotations, and other important facts independently.
Privacy is not the same as clinical confidentiality
Iris is designed as a private consumer service, but no online service can promise absolute confidentiality or security. Iris is not a healthcare provider and does not claim that HIPAA applies. Content may be processed by configured AI, speech, hosting, authentication, storage, security, and support providers; retained under the Privacy Policy; reviewed for limited support, safety, security, fraud-prevention, debugging, or legal purposes; or disclosed when legally required.
Some conversations may qualify as consumer health data under state law. Read the Consumer Health Data Privacy Policy.
Report a concern
To report a harmful response, privacy concern, security issue, or product problem, email app@irislightwithin.com with enough information to investigate. Do not use support email for an emergency.
For general help, visit Support.