Compliance & Safety

Our commitment to a safe and trusted platform.

At dreamz.ai, user safety and regulatory compliance are at the core of our platform. We maintain strict processes to detect, review, and address issues reported by our community.

User Safety First

We prioritize the safety and well-being of our community with proactive monitoring and reporting tools.
Content Review

All reports are reviewed by our dedicated support team to ensure appropriate action is taken.
Data & Privacy

Your information is handled confidentially and used only for safety, compliance, and improvement.
Regulatory Compliance

We follow global compliance standards and collaborate with our banking and payment partners.
Transparency

We believe in transparency and continuously improve our systems based on user feedback.
Accountability

Our team is accountable and committed to resolving issues fairly and promptly.

Our suicide and self-harm prevention protocol

Some people talk to a companion about the hardest things in their life. This is what happens on dreamz.ai when a conversation shows signs of a crisis. We publish it so anyone can hold us to it, and we update this section whenever the protocol changes.

We watch for signs of crisis

Messages are screened for indications of suicidal ideation or self-harm. Detection runs on the conversation itself, not on anything you tell us about yourself, and it is not something you have to opt into.
We show real crisis resources

When those signs appear, we display a referral to trained human help: call or text 988 (the Suicide & Crisis Lifeline) in the United States, and findahelpline.com for the rest of the world. These numbers are fixed text written into the product. They are never generated, chosen, or reworded by the AI, and the notice cannot be dismissed.
We keep a record of every referral

Each time we raise crisis resources in a conversation, we log that referral with a timestamp. The log holds no message content. It exists so we can report how often the protocol was used, including to regulators who require those figures.
We pause the roleplay until the resources are acknowledged

In a high-severity crisis the companion stops the story. The crisis resources take the place of its reply and the conversation does not continue until you acknowledge them.