Last updated: 15 May 2026
Xpirit AI is the data controller for personal data processed through this service. Contact: [email protected]
When you choose "Sign in with Google", Xpirit AI uses Google's OAuth 2.0 service strictly for authentication. The Google API scopes we request are openid, email, and profile — these are Google's standard non-sensitive sign-in scopes. We do not request access to Gmail, Drive, Calendar, Contacts, Photos, YouTube, Maps, Chat, Meet, Tasks, Keep, or any other Google service.
From your Google Account, Xpirit AI accesses only the following fields, returned by Google's ID token and UserInfo endpoint:
sub): an opaque, Google-issued numeric ID that uniquely identifies your Google Account.No other Google user data is requested, accessed, downloaded, copied, scraped, or read.
Xpirit AI uses the Google user data described in Section 3.1 strictly for the following purposes:
sub) or verified email.Google user data is never used for the following:
The fields listed in Section 3.1 are stored only in Xpirit AI's authentication database (PostgreSQL, hosted on infrastructure in the United Kingdom) as part of your user record. Xpirit AI does not share Google user data with any third party except where strictly necessary to provide the service (the database hosting provider, acting as a processor under a data-processing agreement) or where legally required.
Google user data is not shared with any third-party AI or machine-learning provider, including but not limited to OpenAI, Anthropic, Google AI, OpenRouter, Hugging Face, RunPod, or any inference vendor.
Google user data is retained for as long as your Xpirit AI account is active. You can delete your Xpirit AI account at any time from Settings > Account > Delete account, which deletes all stored Google user data within 30 days. You can also revoke Xpirit AI's access at any time from your Google Account permissions page.
Xpirit AI's use and transfer of information received from Google APIs to any other app will adhere to the Google API Services User Data Policy, including the Limited Use requirements.
By using the Service, messages you explicitly rate (thumbs up/down) are collected as training pairs to improve the Xpirit AI model. Only rated interactions are collected — your full conversation history is never bulk-harvested. This is enabled by default as stated in the Terms of Use.
You can disable this at any time via the "Improve AI" toggle in the sidebar or Settings > Privacy. When disabled, no new training pairs are collected. Previously collected pairs are retained unless you request deletion via a Data Subject Request (Art. 17). Legal basis: GDPR Art. 6(1)(b) contract performance (service improvement) with Art. 21 right to object.
To exercise these rights, use the in-app settings or contact [email protected]. We respond within 30 days.
We implement technical measures including: encrypted connections (TLS), scrypt password hashing, immutable audit logs with SHA-256 checksums, rate limiting, IP blocking, content filtering, and admin 2FA via Telegram OTP.
In compliance with EU AI Act Article 52, we disclose: you are interacting with an AI system (Xpirit-Davinci, a custom fine-tuned 7B-parameter language model). The system is classified as limited-risk. It does not perform automated decision-making with legal effects. Technical details available at /api/v1/compliance/transparency.
Your data is processed on servers located in the United Kingdom. For EU users, UK adequacy decision (June 2021) applies. RunPod cloud inference (when active) processes data in the EU/US under standard contractual clauses.
Data Protection Officer: [email protected]
You have the right to lodge a complaint with your national data protection authority (e.g., ICO in the UK, CNIL in France).