Legal

AI Disclosure Statement

Version 1.0 · 11 June 2026 · TheirCall Ltd, Company No. SC887444

TheirCall is built on artificial intelligence, and we believe everyone touched by the service — the family who subscribes and, above all, the relative who receives the calls — should know that plainly. No UK law currently requires this statement. We publish it voluntarily, because a service built on daily conversations with elderly people should hold itself to the standard of honesty it would want for its own family.

1. Where AI is used in TheirCall

• The daily calls themselves are conducted by an AI voice — a warm, consistent companion persona named Francis or Molly, chosen by the family at sign-up. There is no human caller.
• The wellbeing summaries you receive after each call are written by AI from the conversation.
• The service uses AI to gradually learn your relative's conversational preferences — the topics they enjoy, how they like to be addressed — so the calls become more personal over time.

Humans run the company, design the service, and oversee how the AI behaves; the AI conducts the conversations and writes the summaries.

2. Who is told, and when

Families: you are told that calls are made by an AI voice on this website, and again — explicitly, with its own individual acknowledgement — during sign-up, before any payment is taken.

Your relative: before the daily service ever begins, the very first call your relative receives is an introductory call in which the service is explained to them plainly — including that the caller is an AI — and they are asked directly for two separate consents: to receive the calls, and for summaries to be shared with you. The daily calls begin only if they give both. No one consents on their behalf.

3. How the AI behaves about its own nature

On the calls themselves, the AI follows a firm, simple rule:

• It never claims or implies that it is human, and never denies being an AI.
• If your relative asks directly — "are you a real person?" — it answers honestly and warmly.
• If your relative shows genuine confusion about what or who they are talking to — for example, mistaking the AI for a real family member — it gently corrects them.
• It never invents a human life for itself.
• On ordinary, friendly small talk, it simply stays the warm companion it is. It does not interrupt every conversation with a robotic announcement that it is an AI, because for someone who already understands — and has consented — that adds nothing but coldness to a call whose whole purpose is warmth.

An honest note on the limit of this approach: gentle correction relies on confusion being noticeable in the conversation. That is one of the reasons the disclosure is made so explicitly at the start — to the family at sign-up and to the relative on the introductory call — and why we encourage families to talk to their relative about the service. If you ever feel your relative no longer understands what TheirCall is, contact us at hello@theircall.co.uk and talk to us.

4. What the AI will never do

The AI never gives medical advice, diagnosis, or treatment recommendations. It never claims to be human. It never makes promises on TheirCall's behalf. If it ever behaves outside these boundaries, we treat that as a serious incident: the behaviour is investigated, the AI's instructions are audited and rolled back where needed, and affected families are told.

5. The records we keep

The AI's instructions — including the exact words used to explain the service and ask for consent on the introductory call — are formally version-controlled. Every change is numbered, dated, and recorded with its reasoning, and previous versions are archived.

We go one step further with consent: every consent your relative gives is permanently recorded together with the version of the disclosure wording that was in force at that moment. That record cannot be altered. It means we can always show not only that consent was given, but exactly what your relative was told when they gave it.

6. The standard we align to

The UK does not yet have AI-specific legislation of this kind. TheirCall voluntarily aligns with the transparency principles of the EU Artificial Intelligence Act for services of this nature — the core of which is simple: people should know when they are talking to an AI. We adopt that principle without qualification, and this statement, the sign-up disclosures, and the introductory call are how we meet it.

7. Questions

If anything here is unclear, or you want to understand more about how the AI works, contact us at hello@theircall.co.uk — a human will answer.

Version history

v1.0 — 11 June 2026 — First published version.