AI & Health-Data Disclosure

Last updated: October 12, 2026

This page explains how AI answers your messages on PlanGains, what it can see, and how safety works. It sits alongside our Privacy Policy and Terms of Service.

How replies work

When you message a Creator, an AI assistant writes the reply on the Creator’s behalf, based on the Creator’s own materials. Your chat shows that replies come from the Creator’s AI assistant. Routine replies are sent automatically, so you get an answer right away. These wait for your Creator to review first:

  • a reply suggesting a weight the AI can’t check against your workout screen;
  • a request to change your program;
  • every reply, if your Creator has chosen to review everything.

Your Creator sees the whole conversation, including what the AI first wrote and what was actually sent, and can retract a reply. If a reply is waiting for your Creator, your chat says so.

This is not medical advice

Replies — whether AI-written or from your Creator — are general fitness guidance only and not a substitute for professional medical advice, diagnosis, or treatment. Talk to a qualified healthcare provider about your health. In an emergency, call your local emergency number.

What the AI sees

To answer you, the AI receives:

  • your message, your recent conversation with that Creator, and a running summary of it;
  • the Creator’s own materials — their philosophy, coaching rules and programs;
  • your program and today’s workout, including the weights your workout screen suggests;
  • the sets you’ve logged on that Creator’s programs, and your bests;
  • your levels on that Creator’s paths;
  • your profile basics: goal, experience, height, weight and units;
  • injuries you’ve reported and which lifts are on hold because of them, and starting weights you’ve given.

The AI does not see photos or videos. A form-check video goes to your Creator to watch personally.

Across Creators

If you follow more than one Creator, each Creator’s AI sees only your conversation and training with that Creator, with three exceptions that keep your training safe and consistent:

  • the weights suggested on your workout screen and in replies are worked out from all your logged training;
  • a starting weight you give for a lift in any Creator’s chat is used for that lift with all your Creators;
  • an injury you report holds the suggested weight on the affected lifts with all your Creators. A Creator you didn’t tell — and their AI — see only that the lift is on hold, not why, unless you choose to share it.

Injuries and pain

If you mention an injury or pain, the reply is written not to diagnose or advise on medication, and it carries a short note that it isn’t medical advice. Your Creator chooses one of two settings for how these replies are sent:

  • Reviewed first: you get a fixed safety message straight away, and the AI’s answer waits for your Creator to review and send.
  • Answered straight away: the AI’s answer is sent at once with the note, and your Creator is told about it.

Either way, PlanGains asks whether to hold the suggested weight on the lifts it might affect, and only holds it if you say yes.

Emergencies

A message that suggests a medical emergency or a personal crisis gets an immediate fixed safety message instead of an AI reply — telling you to stop and call your local emergency number, or giving crisis-line contacts — and your Creator is alerted urgently. The safety message is fixed text, not written by the AI, and a list of emergency phrases triggers it even if the AI model is unavailable.

If AI replies are switched off in your chat — your Creator has paused them, or you’ve asked to talk to your Creator directly — nothing automated runs, including these safety messages: your messages go straight to your Creator. In an emergency, always call your local emergency number rather than relying on a chat.

Reply allowance

Each subscription includes instant AI replies up to an allowance per billing period. If you reach it, AI replies pause for that subscription until your next billing date and your Creator answers you personally. The safety checks and messages above keep running.

Which companies process the data

Replies and safety checks run on Google Cloud (Vertex AI, Gemini models). When Google is slow or unavailable, a reply is written by Anthropic (Claude) instead. Both are used under commercial terms that do not allow your data to be used to train their models, and your data is sent only to produce that reply or check. Google may process a request in any of its regions worldwide; Anthropic processes in the United States. Where everything else is processed is in our Privacy Policy.

Your choices

  • You don’t have to share medical details in chat — share only what you’re comfortable with.
  • You can ask to talk to your Creator directly at any time.
  • For medical concerns, contact a healthcare professional.
  • You can access or delete your data as described in our Privacy Policy.