§ LEGAL · ODYSSEY TECH SAS

AI Use Disclosure

How Odyssey uses artificial intelligence tools, our data protection commitments, and EU AI Act compliance. Last updated: 27 May 2026.

Compliance: EU AI Act (EU 2024/1689) · GDPR Article 22 · CNIL 2026 guidelines

§ AVAILABLE IN FRENCH · DISPONIBLE EN FRANÇAIS

A bilingual French/English version of this policy is available as a PDF document.

↓ DOWNLOAD BILINGUAL PDF (FR/EN)

Preamble

As a firm specializing in artificial intelligence strategy, Odyssey uses generative AI tools in its daily operations. This transparency is essential to our trusted relationship with our clients.

This document, complementary to our privacy policy, details our use of AI, our data protection commitments and our obligations under the EU AI Act (EU Regulation 2024/1689).

1. Our philosophy

Odyssey conceives artificial intelligence as a tool augmenting human analysis, not replacing it. Our fundamental convictions:

• AI accelerates certain analysis tasks, but never replaces professional judgment.

• Work delivered to our clients is systematically validated by a partner or senior consultant.

• Sensitive data is never submitted to consumer interfaces.

• Confidentiality takes precedence over productivity.

2. AI tools used

Odyssey uses the following generative AI tools for professional purposes:

• Claude (Anthropic, USA) — primary use, analysis, drafting, structuring. API with Zero Data Retention. No training on customer data.

• ChatGPT (OpenAI, USA) — occasional use. Enterprise / Team accounts, no training opt-out activated.

• Gemini (Google, USA) — occasional use. Google Workspace integration, no training enabled.

• Grok (xAI, USA) — occasional use. Professional accounts with no training settings.

All these tools are used as commercial services with contractual agreements with their publishers.

3. Commitments regarding your data

3.1 No model training

Odyssey formally commits to never use the data of its clients, prospects or partners to train artificial intelligence models, and to authorize no third-party publisher to do so with this data.

This commitment is implemented through systematic activation of "no training" options in professional accounts, priority use of APIs in Zero Data Retention (ZDR) mode, regular audit of our internal practices, and contractual review of agreements with AI publishers.

3.2 Use segregation

For sensitive engagements (M&A, due diligence, confidential financial data): exclusive use of Claude API (Anthropic) in ZDR mode, no input to consumer interfaces (Claude.ai, ChatGPT.com, etc.), internal prompt logs for traceability.

For internal uses (research, article drafting, training preparation): standard use via professional interfaces, no identifiable client data.

4. EU AI Act compliance

4.1 Classification of our uses

EU Regulation 2024/1689 on Artificial Intelligence ("EU AI Act"), which entered into force on August 1, 2024, classifies AI systems according to their risk level. Odyssey operates as a professional user ("deployer") of AI systems, not as a provider.

No unacceptable risk uses (Article 5 AI Act). No high-risk uses (Annex III AI Act). Uses falling under the "limited risk" category (Article 50).

4.2 Compliance obligations

Transparency (Article 50 AI Act): informing our clients about the use of AI tools in preparing deliverables, clear identification of AI-generated or AI-modified content in our productions, and this public policy.

Human oversight: systematic validation by a qualified human, ability to refuse or modify AI suggestions, staff training on risks and best practices.

AI literacy (Article 4 AI Act): continuous staff training on AI capabilities and limits, active regulatory and technological monitoring.

5. No automated decisions

In accordance with Article 22 of the GDPR, Odyssey makes no decisions producing legal effects or significantly affecting a person, based solely on automated processing.

Concretely: our strategic recommendations are always formulated and validated by humans. AI proposes, humans decide. No automated scoring of candidates, clients or investment targets. No automated profiling of natural persons.

6. Your rights

6.1 Right to information

You have the right to know if and how AI was used in preparing a deliverable that concerns you. Upon request, we will indicate the tools used and their contribution.

6.2 Right to object to AI use

You may request that Odyssey handle your engagement or request without recourse to generative AI tools. This request must be made in writing (contact@odyssey.tech), will be studied based on technical and operational feasibility, may affect the timeline and pricing of our intervention, and will receive a reasoned response within 15 business days.

7. Policy evolution

The AI field evolves rapidly, both technologically and regulatorily. This policy will be updated at least every six months, and more frequently if necessary to reflect evolutions of the EU AI Act and its guidelines, introduction of new AI tools in our practices, evolutions of publisher commitments (Anthropic, OpenAI, Google, xAI), and recommendations from the European Data Protection Board (EDPB).

8. Contact

For any question regarding our AI use, your rights, or for commercial inquiries: contact@odyssey.tech.

§ CONTACT

Questions about our AI practices?

contact@odyssey.tech — Single address for all inquiries, including exercising your GDPR rights.