Privacy policy
AI ARTEDUSA privacy policy: processing of personal and voice data, GDPR compliance, sub-processors, retention periods.
ARTICLE 1 - DATA CONTROLLER
AI ARTEDUSA is the data controller for data related to the client relationship: account, billing, and the institution's staff use of the portal — including voice dictation in the carnet assistant, whose audio belongs to that relationship and not to caller data. For the personal data of callers and end users processed via the Service (audio recordings, transcripts, summaries), the client institution — museum, library, theatre, gallery, or the local authority it falls under — is the data controller, and AI ARTEDUSA acts as a processor within the meaning of Article 28 of the GDPR, on behalf of and on the instructions of the controller. A data processing agreement (DPA) is made available to clients. Contact: support@artedusa.com.
ARTICLE 2 - DATA COLLECTED
As part of the operation of the Service, the following categories of data are collected and processed:
- Identification data: first name, last name, email address, phone number, company name
- Billing data: postal address, SIRET number, banking information (managed by Stripe)
- Voice data: audio recordings of telephone conversations between callers and the AI bot
- Transcript data: text from the voice-to-text conversion of conversations
- Usage data: login logs, call statistics, dashboard navigation data
- Knowledge base data: documents uploaded by the User
ARTICLE 3 - PURPOSES AND LEGAL BASES
Data is processed for the following purposes:
Performance of the contract: Operation of the Service (receiving and processing calls, transcription, summary), user account management, billing.
Legitimate interest: Service improvement, anonymised statistical analysis, fraud and abuse prevention.
Legal obligation: Retention of invoices and accounting data in accordance with applicable legislation.
ARTICLE 4 - SUB-PROCESSORS AND DATA TRANSFERS
Data is processed by the following sub-processors, all contractually bound to comply with the GDPR:
Voice processing for telephone calls (speech recognition and synthesis), semantic search, and data storage are provided on a European infrastructure controlled by AI ARTEDUSA, without recourse to a third-party sub-processor: this data does not leave the European Union. The portal's spoken assistant is the exception: dictation and read-aloud, available to the client institution's own staff, are provided by Mistral AI (France, European Union). The audio is never kept there — it is transcribed and then discarded; only the text remains, under the retention periods set out in Article 5.
The sub-processors listed above are established in the European Union, with two exceptions, both governed by the European Commission's Standard Contractual Clauses (decision 2021/914). First, for card payments by private clients (B2B) only, Stripe — as an international group — may use processing in the United States. Second, a private institution (B2B) may choose Anthropic as its language-model house in its settings: the TEXT of its conversations is then processed in the United States. That choice is never active by default, it is taken with an explicit, timestamped consent, it can be reversed at any time, and public-sector institutions have no access to it.
- OVHcloud (French group, hosting in the European Union): hosting of the application infrastructure and data (servers, PostgreSQL database, Redis cache, Qdrant vector database), telephony connection (SIP trunk), and sending of transactional emails (SMTP)
- Mistral AI (France, EU): default language model and embedding computation for conversation processing and augmented search; speech recognition and synthesis for the portal's spoken assistant, with the audio not retained
- Stripe Payments Europe (Ireland, EU): card payment processing, for private clients (B2B) only
- Anthropic PBC (United States): language model, ONLY if your institution has chosen it in its settings — never by default, and never for a public-sector institution. Neither embedding computation nor speech recognition and synthesis are entrusted to it: they remain in the European Union regardless of that choice
ARTICLE 5 - RETENTION PERIODS
Data is retained for the following periods:
- Account data: retained throughout the use of the Service, then deleted upon account closure. Upon erasure request (right to erasure, Art. 17 GDPR), it is deleted within the shortest possible time.
- Voice data (audio recordings): 90 days after the call, unless an early deletion request is made
- Transcripts and summaries: 12 months after the call
- Billing data: 10 years in accordance with accounting obligations
- Login logs: 12 months
ARTICLE 6 - RIGHTS OF DATA SUBJECTS
In accordance with the General Data Protection Regulation (GDPR), you have the following rights:
- Right of access: Obtain a copy of your personal data
- Right of rectification: Correct inaccurate data
- Right to erasure: Request the deletion of your data
- Right to data portability: Receive your data in a structured format
- Right to object: Object to the processing of your data
- Right to restriction: Request restriction of processing