AI Notice

Some conversations on this site — including the AI architect intake and inbound channels — may be handled by autonomous systems before being escalated to a human operator. Routing happens silently in the background.

AI outputs can be incomplete or wrong. Do not rely on AI-generated content as the sole basis for legal, medical, financial, or safety-critical decisions. Verify anything that matters.

We do not use your private conversations to train public third-party models. Internal model improvements happen on isolated infrastructure with access controls.

You can ask to speak with a human architect at any time. The operator layer responds asynchronously within one business day.

If an AI response is wrong, harmful, or unsafe, tell CRABX AI. We treat this as a priority signal.