LAUNCHING 5 NOVEMBER 2026 YOUR DATA STAYS ON YOUR PHONE BUILT IN THE EU

AI transparency

Whether The Veda Circle uses AI, and exactly what that means for you.

AI Transparency Policy · Effective Date: 23 October 2026 · Version 1.1

1. Current Position: No AI System in the App

As at the effective date of this policy, the App's cycle predictions and all other in-App insights are calculated using fixed, rule-based logic (calendar and date arithmetic, averages, trends, and simple pattern rules) applied to the data you enter on your device. They do not use machine learning or generative AI. There is no chatbot, no generated content, and no AI-driven decision-making in the App. On that basis:

  • GDPR Article 22 (automated decision-making) does not apply. Cycle prediction involves automated processing of personal data on your device and may be regarded as profiling within the meaning of GDPR Article 4(4), but it produces no decision based solely on automated processing that has legal effects on you or similarly significantly affects you. This is also stated in Section 7 of our Privacy Policy. We will reassess this position if the App ever surfaces insights that feed into decisions with legal or similarly significant effects, including (but not limited to) insights relating to conception, termination of pregnancy, insurance, employment, or integration with reproductive healthcare providers.
  • The EU AI Act's transparency obligations for AI systems (Article 50, for example disclosing that a person is interacting with an AI system and labelling AI-generated content) do not apply, because the App does not contain an AI system as defined by that Regulation.

We keep an internal technical record listing the parts of the App that produce insights and the formulas and rules they use. We update it whenever insight-related code changes. It is the evidence behind the statements in this Section.

2. What Counts as AI for This Policy

  • Not AI, for this policy: fixed calendar and date arithmetic (for example, predicting a next cycle date from your logged past dates using a set formula); averages, trends, and comparisons of your own data; simple rule-based logic (for example, if a symptom is logged several times, show a suggestion or a link to a tool); and static, human-written educational content.
  • AI or machine learning, for this policy: any feature of the App whose behaviour is learned from data (a trained model) rather than explicitly programmed rules, including machine learning that is part of the App, pattern recognition across your history that goes beyond simple thresholds, any chatbot or conversational feature, or any feature that calls a third-party AI or large language model service.

If it is unclear which category a feature falls into, we treat it as AI for the purposes of this policy and apply the commitments in Section 5 before it is released.

3. Features That Use Your Phone's Own Technology

Two features use technology that is built into your phone's operating system and is provided by the phone's maker (Apple or Google), not by us:

  • Voice input: if you tap the button to speak instead of typing, your phone's own speech recognition turns your voice into text. The App asks the phone to do this on the device, and shows the button only on phones that can. Your voice is not recorded or uploaded by the App, and we never receive it.
  • Read-aloud: your phone's built-in text-to-speech voice reads the screen to you. Nothing is sent anywhere to make this happen.

Speech recognition and text-to-speech on a phone are themselves built using machine learning by the phone's maker. They are not part of the App's own systems, they run on your device, and we do not train, control, or receive anything from them. We mention them so that this policy is complete. If we add integration with a voice assistant (such as Siri) in future, we will update this policy and our Privacy Policy before it is offered.

4. How AI Tools Were Used to Build the App

AI-assisted software tools were used by the developer in building the App and in preparing some translations. These tools are not part of the App, do not run inside it, and never receive your data. Some translations have not yet been reviewed by professional translators; see our Terms and Conditions.

5. Commitments If AI Features Are Added

If any AI feature is added to the App in future, before it is released we will:

  • Update this policy, the Privacy Policy, and the Terms and Conditions to describe the feature accurately: what it does, what data it uses (and whether that requires any change to the App's on-device design, which would itself be a major change needing its own review), and its limitations.
  • Where the feature involves a chatbot, generated content, or anything a user could reasonably mistake for a human or for verified medical fact, clearly tell users that they are interacting with an AI system, consistent with EU AI Act Article 50 and equivalent expectations elsewhere.
  • Where the feature would process health data through any third-party AI service, treat that provider as a processor under a data processing agreement, assess whether the data is de-identified, and complete a data protection impact assessment. This would be a significant change to the App's current design, under which what you log does not leave your device, and would be treated accordingly.
  • Not allow any AI feature to produce output that could be mistaken for medical advice or diagnosis, consistent with Section 5 (Not Medical Advice) of the Terms and Conditions. Any AI-generated insight would carry the same disclaimers as the rest of the App.
  • Give users a basic explanation of what the feature does with their input and, where feasible, a way to opt out while still using the rest of the App.

6. Human Oversight

Sarah Lesley Hughes is responsible for reviewing any AI feature before release and for keeping this policy accurate. No AI feature used by the App, now or if one is added later, is intended to operate without the ability for a human to review, override, or disable its output.

7. Review

We review this policy whenever a new feature is proposed that might involve AI, and at least once a year, because AI regulation (in particular the phased obligations of the EU AI Act) continues to develop. Date of last review: 20 October 2026. Next review due: 20 October 2027.

8. Contact

Questions about this policy can be sent to hello@thevedacircle.com.

Version History

  • Version 1.0, 2 October 2026: Initial issue Approved by Sarah Lesley Hughes.
  • Version 1.1, 23 October 2026: Rewritten for publication: internal instructions and placeholders removed; contact address corrected; section numbering and change log corrected; on-device speech recognition and read-aloud disclosed; AI-assisted development and translation disclosure added; cross-reference to Privacy Policy Section 7 added; approver name made consistent Approved by Sarah Lesley Hughes.