Safety · Policy

AI Transparency & Responsible AI Policy

AI should expand human experiences, not deceive people about what is real.

Last updated · September 2026

Overview

Artificial intelligence is fundamental to many of the experiences we build at Imperfect Lab.

Our products may include AI personalities, AI-generated conversations, synthetic media and other AI-powered experiences.

As these technologies become increasingly realistic, we believe transparency becomes increasingly important.

This policy describes the principles that guide how we use AI across our products.

1. AI should be identifiable

Users should be able to understand when they are interacting with an AI personality rather than a real person.

We design our products to distinguish AI profiles and experiences from real human creators.

AI personalities must not be deceptively represented as real human beings.

2. AI-generated experiences

Our products may use artificial intelligence to generate or assist with:

Conversations.
Images.
Video.
Voice.
Personality and character interactions.
Recommendations and personalization.
Other interactive experiences.

AI-generated outputs may sometimes be inaccurate, unexpected or unrealistic.

Users should not treat AI-generated information as professional, medical, legal, financial or other expert advice.

3. Human + AI

We don't believe AI needs to replace human interaction.

Our vision is to create products where real people and AI personalities can coexist while remaining clearly distinguishable.

Human creators bring real human interaction.

AI personalities create a different kind of experience powered by technology.

Users should be able to understand that difference.

4. Synthetic media

AI can generate increasingly realistic images, video and audio.

We use product design, labeling and other contextual signals where appropriate to help users understand when they are experiencing synthetic or AI-generated content.

We prohibit the use of our systems to create certain harmful or deceptive synthetic media, including prohibited non-consensual sexual deepfakes and sexualized representations of minors.

5. Child safety

AI must never be used to circumvent our child-safety standards.

We prohibit the use of our AI systems to generate, request or facilitate sexual content involving or representing minors.

This prohibition applies regardless of whether the depicted person is real, fictional or entirely synthetic.

See our Child Safety & CSAM Policy for more information.

6. Safety by design

We use safeguards designed to reduce the generation and distribution of prohibited content.

These may include technical controls, model-level safeguards, moderation systems, product restrictions and human review.

No safety system is perfect.

We continuously evaluate and improve our safeguards as AI technologies and associated risks evolve.

7. Privacy

AI systems can require data to provide personalized experiences.

We aim to design these experiences in accordance with applicable privacy and data protection requirements.

For information about what data is collected, why it is processed and how long it is retained, users should review the Privacy Policy applicable to the relevant product.

8. Responsible innovation

AI is changing quickly.

Policies and safeguards must evolve with it.

We continuously evaluate emerging technologies, risks, regulations and industry practices as we develop new AI-powered experiences.

Our principle is simple:

People should understand when they're interacting with AI, remain in control of their experience, and be protected by clear boundaries around harmful uses.

Our vision

We believe the future of companionship is hybrid.

Human connection and artificial intelligence can create fundamentally different experiences, and both can exist in the same digital world.

Our goal is not to blur the distinction between them.

It's to explore what becomes possible when both are brought together responsibly.