This document is published in English. Translations for other markets will be published once legally reviewed.
This page explains, in plain language, how MyPA uses artificial intelligence, how your data is used, and what AI can and cannot do. It should be read with our Terms of Service and Privacy Policy. A link to this page appears in the website footer, the app's Settings > About AI menu, the chat screen and checkout.
1. MyPA uses AI-generated content
MyPA is an artificial intelligence (AI) assistant. Most of what you see in MyPA is created by AI, including:
- chat responses, answers, summaries, translations and drafts;
- search results summaries, recommendations and comparisons;
- images, voices, audio and video;
- Creator AI Personas (AI characters configured with a creator's permission to talk in their style or voice); and
- messages, bookings and other actions MyPA prepares or carries out for you.
You are talking to an AI unless we clearly tell you otherwise. When a real person is involved, such as an Ask the Experts professional or a human support agent, the screen will show their name and a "Human" or "Expert" label.
We label AI content so you can recognise it. Key labels include:
| Label | What it means |
|---|---|
| AI | Generated by AI. Not checked by a person before you saw it. |
| AI image / AI audio / AI video | Media created or significantly altered by AI. Downloads also carry machine-readable markings where technically possible. |
| AI Persona | You are chatting with an AI configured in the style of a creator or character, not the real person. |
| AI + Human reviewed | Generated by AI and checked by a qualified person before release. |
| Human / Expert | Written or said by a named person. |
| Sponsored / Ad | Paid placement or commercially influenced content. |
Full details are in our AI Content Labelling System Policy.
2. How AI works in MyPA
MyPA chooses the best way to handle each task:
- On your device (Local Lite / Local Smart): smaller AI models run on your phone or computer where possible, so your information does not need to leave your device for that step.
- In the cloud (Cloud Smart): for more complex tasks, the minimum information needed is sent securely to cloud AI models run by us or by carefully selected AI providers acting on our instructions.
- Your own AI provider (where supported): if you connect your own AI account or key, information is sent to that provider under its terms.
Web searches, bookings, purchases and messages always involve online services, even if the language processing happens on your device.
3. How your data is used
We use your data to provide the service you ask for. For example, to answer your question, remember your preferences (if you turn memory on), connect to your email or calendar (if you link them) or complete a booking you have approved.
What we do:
- process your prompts, files and connected-account data only as needed for the task you request;
- keep your personal memory for your benefit, which you can view, edit or delete in settings;
- share the minimum details needed with a third party when you ask MyPA to act (for example your name with a restaurant you book);
- use security and fraud-prevention data to protect you and the Service; and
- use aggregated, de-identified usage statistics to improve reliability and quality.
What we do not do:
- we do not sell your private conversations, email content or personal data to advertisers;
- we do not allow our third-party AI providers to use your data to train their general models;
- we do not use your private conversations, emails or files to train general-purpose AI models;
- we do not use Gmail or Google Workspace data, Apple VPN data, health data or payment credentials for advertising or MyPA Insights; and
- we do not enrol you in MyPA Insights (our optional rewards-for-data programme) unless you actively opt in. You can leave at any time.
Payments: card payments are handled by our payment provider, Trust Payments. MyPA never sees or stores your full card number.
Your choices: you can turn memory, personalisation, connected accounts, cloud processing preferences, MyPA Insights and marketing on or off in Settings > Privacy, and download or delete your data. See our Privacy Policy for full details, retention periods and your legal rights, or contact dpo@mypa.today.
4. Limitations of AI – please read
AI is powerful but imperfect. You should not rely on MyPA as your only source for anything important.
- AI can "hallucinate". It may produce information that sounds confident and convincing but is wrong, incomplete or entirely made up, including false facts, names, figures, quotes, links, references, legal cases or product details.
- AI can be out of date. Information may not reflect recent events, prices, availability, laws or opening hours.
- AI can misunderstand. It may misread your request, a document or an image, or mix up people, dates or amounts.
- AI can be biased or inappropriate. Despite safeguards, outputs may reflect biases or be unsuitable for your circumstances.
- AI is not a professional. AI answers about health, law, money, tax, insurance or safety are general information, not professional advice. Use Ask the Experts or consult a qualified professional for decisions that matter.
- AI is not an emergency service. If someone is in danger, contact your local emergency services immediately.
- AI can make mistakes when acting for you. That is why MyPA asks you to confirm purchases, bookings and other important actions. Always check the details on the confirmation screen.
- AI Personas are not real people. A Creator AI Persona does not know you personally and its statements are not the creator's own words or advice.
- Recommendations may be influenced commercially. Sponsored results are labelled. Always compare options that matter to you.
What we do about it: we test models before release, use safety filters, ground answers in search results or your own data where possible, show sources where available, require confirmation for important actions, route high-risk matters to human review, and monitor reported errors. See our Human Oversight Process Policy and AI Ethical Framework Policy.
5. Reporting a problem
If you see an AI output that is wrong, harmful, offensive, or not labelled correctly:
- tap the flag / thumbs-down icon on the response; or
- email help@mypa.today (or dpo@mypa.today for privacy concerns).
Reports are reviewed by our team. You can also ask for a person to review any decision that significantly affects you, such as an account restriction.
6. Our policies
- AI Ethical Framework Policy
- Human Oversight Process Policy
- AI Content Labelling System Policy
- Terms of Service and End User Licence Agreement
- Privacy Policy
- Payments, Subscriptions, AI Tokens and Refund Policy