Yes, always, and on AccessMe it is not optional: every twin page states that the twin is an AI representation trained on its creator's content, not the person themselves. Disclosure protects your audience from being misled, protects you from a promise you did not make, and protects the twin's usefulness, because people use an AI well only when they know it is one. Honesty about the twin is what lets the real relationship survive it.
Why people hide it
A good twin is convincing. Someone building one might reason that the audience will engage more if they believe they are talking to the person. For a while, they might. Then one of them finds out, and the discovery does not read as "clever technology"; it reads as "you let me pour my heart out to a machine while letting me think it was you." Every conversation before it is retroactively poisoned.
What disclosure protects
- Your audience. They can decide what to share, how much weight to give an answer, and when to ask for the real you. People who know they are talking to an AI representation ask differently and more usefully.
- You. A twin that people believe is you is making promises in your name. Every answer becomes a commitment you did not make. Labeling the twin keeps the line clear: this is my point of view, carried by an AI; I am behind it, not inside it.
- The twin itself. A labeled twin can be what it is: a front door to you, reliable on what you have said, honest about what you have not. An unlabeled one has to keep up a pretense it will eventually fail.
How AccessMe handles it
Every twin page carries an AI disclosure. Every twin demo on this site does too. The footer of every page says it: a twin is an AI representation trained on its creator's content, not the person themselves. A creator cannot turn this off, and we think that is a feature. It also means a creator never has to have the awkward conversation; the page has already had it.
Disclosure is not a disclaimer
A disclaimer is small print to limit liability. Disclosure is a plain statement that lets people trust what they are told. The label on a twin is the second kind. It does not say "do not trust this"; it says "this is Dana's thinking, carried by an AI, and Dana can see what you share with it." That is more trustworthy than most things people read online, precisely because it says what it is.
The rest of honest practice
- Let people choose private mode when they do not want you to read a conversation.
- Do not let the twin promise that you will follow up. It does not notify you, and you decide.
- Keep the twin within what you would say in public, so it never claims a view you do not hold.
- Say, in your own voice somewhere, why you built it. Most audiences are glad: it means they can reach your thinking at midnight without feeling they have imposed on you.
Common questions
Can I remove the AI label from my twin?
No. Every AccessMe twin page states that the twin is an AI representation trained on its creator's content, not the person themselves. The label is part of what makes a twin trustworthy.
Will people engage less if they know it is an AI?
People who know what they are talking to tend to ask better questions, share appropriately, and trust the answers for what they are. Being misled is what ends engagement.
Does the twin tell people that I will read their conversation?
The audience member chooses normal or private mode on paid plans. In normal mode you can read and follow up; in private mode you cannot. The twin never promises that you will reply.