Odiseo
Español (ES)Back to home

AI Transparency Notice

Last updated: 10 August 2026

This notice is published by Nexus Generated FZ-LLC, a free zone limited liability company licensed by the Ras Al Khaimah Economic Zone (RAKEZ), License No. 5034603, with registered address at VUET1700, Compass Building - Al Hulaila, Al Hulaila Industrial Zone-FZ, Ras Al Khaimah, United Arab Emirates ("Nexus Generated", "we", "us").

Nexus Generated operates Odiseo (odiseo.ai), a platform that lets businesses deploy AI-powered digital employees to communicate with their website visitors and customers through chat, voice, WhatsApp, and email.

This notice explains, in plain language, when you are interacting with artificial intelligence on Odiseo or through an Odiseo-powered agent, what that AI can and cannot do, and how the technology works at a level that is useful to you. It is written to satisfy the transparency obligations of Article 50 of the EU Artificial Intelligence Act (Regulation (EU) 2024/1689), in force since 2 August 2026, and it applies wherever you encounter an Odiseo agent, regardless of your location.

1. You are talking to an AI — and we tell you so

Every conversational agent operated through the Odiseo platform is an artificial intelligence system. It is not a human being, and we do not present it as one.

We disclose this at the start of the interaction, in the channel you are using:

  • Chat (website widget): the conversation opens with a clear statement that you are speaking with an AI assistant. This disclosure appears before the substance of the conversation begins.
  • Email: messages generated by an Odiseo agent identify themselves as coming from an AI assistant. This identification is appended by our servers; the AI cannot edit or remove it.
  • Voice (live calls) and WhatsApp: disclosure for these channels is being brought to the same standard as chat and email.

If a business deploys an Odiseo agent on its own website or channels, that agent represents the business — but it remains an AI system, and the disclosure travels with it. The disclosure is part of the platform, not an optional setting a business can quietly remove.

2. What the AI can do

An Odiseo agent is designed to act as a digital employee for the business that deploys it. Within that role, it can:

  • Answer questions using the knowledge base the business has provided (documents, product information, policies).
  • Hold natural conversations in text and voice, in your language.
  • Qualify and record sales interest, capturing information you choose to share into the business's customer records.
  • Perform supervised actions the business has enabled, such as creating support tickets, scheduling meetings, or logging tasks.
  • Share a payment link belonging to the business's own payment account, when the business has configured one. Odiseo itself does not process or hold those payments.

3. What the AI cannot and will not do

These are deliberate, enforced limits — not aspirations:

  • No professional advice. Odiseo agents do not provide financial, legal, or medical advice. This restriction is enforced by dedicated safeguards built into the platform (two independent layers) that decline requests for financial, legal, or medical advice, so that the limitation does not depend on the AI "remembering" its instructions. If you ask for professional advice, the agent will decline and, where appropriate, suggest you consult a qualified professional.
  • No impersonation of humans. The agent will not claim to be a person.
  • No guaranteed answers. Like all current AI language technology, the agent can make mistakes, misunderstand a question, or produce an answer that is incomplete or incorrect (see Section 6).
  • No unsupervised spending or unbounded behavior. Each deployment operates under per-client usage caps, and businesses' data environments are strictly isolated from one another.

4. A human is available

AI does not replace the ability to reach a person. Odiseo agents support escalation to a human: when a conversation needs human attention — because you ask for it, or because the situation calls for it — the agent can hand the conversation off to the business's team, and the business is able to take over the conversation directly. Once a human has taken over, the AI does not talk over them.

If you are interacting with an Odiseo agent deployed by a business, that business remains your point of contact for the substance of your inquiry. For questions about the AI system itself, you can contact us (Section 9).

5. The technology behind Odiseo

Odiseo is built on foundation models and speech technology from established providers. Currently, our conversational AI relies on:

  • Anthropic — large language models powering the agents' understanding and responses;
  • Google — language models used for auxiliary processing tasks;
  • ElevenLabs — voice technology for live speech conversations.

We do not disclose internal system architecture, prompts, or configurations, as these are proprietary and their disclosure would not add meaningful transparency for users. What matters for you: the text and speech you receive from an Odiseo agent is machine-generated by these AI systems, orchestrated by our platform.

A full list of the service providers that process data on the platform is available in our Privacy Policy and Data Processing Agreement.

6. Accuracy: AI can be wrong

AI language models generate responses based on patterns in data. They are highly capable, but they are not infallible:

  • An agent's answer may be inaccurate, outdated, or incomplete, even when it sounds confident.
  • Responses are generated from the knowledge base the deploying business provided; if that material is wrong or stale, answers may be too.
  • Do not rely on an AI response alone for decisions that matter. Verify important information with the business directly or with a qualified professional.

The platform includes safeguards that reduce — but cannot eliminate — these risks, including an automated accuracy safeguard that programmatically corrects certain kinds of overclaiming by the agent. No AI safeguard is perfect, and we do not represent otherwise.

7. AI-generated content is labeled as such

Content produced by an Odiseo agent is presented as coming from an assistant, not from a named human:

  • In chat, AI replies appear under the assistant's identity within a conversation that has been disclosed as AI-driven from the start.
  • In email, AI-generated messages identify their AI origin; the identification is appended server-side.
  • Voice and WhatsApp: being brought to the same disclosure standard as chat and email.

Odiseo agents do not generate content designed to pass as human-authored, and the platform does not offer tools to strip these disclosures.

Marking of synthetic content (Article 50(2) of the EU AI Act). Beyond telling you that you are talking to an AI, the AI Act also addresses the machine-readable marking of synthetic content itself. Odiseo's outputs include AI-generated text (chat, email) and AI-generated audio (live voice synthesized through our voice provider). Conversational AI content is disclosed as such in context. We are evaluating machine-readable marking of synthetic audio as the technical standards under Article 50(2) mature.

8. Your data in AI conversations

This notice is about transparency; data protection is covered in detail in our Privacy Policy. Two points belong here because they concern the AI directly:

  • Your conversations are not used to train AI models. We do not use the content of client conversations to train foundation models, and we select providers consistent with that commitment.
  • Who is responsible for your data depends on where you met the agent. If you are talking to an Odiseo agent on a business's website or channels, that business is the data controller and we process data on its behalf. If you are on odiseo.ai itself, Nexus Generated is the controller. The Privacy Policy explains both cases.

9. Regulatory basis and contact

This notice implements the transparency obligations of Article 50 of the EU AI Act for AI systems intended to interact directly with natural persons: persons are informed that they are interacting with an AI system, in a clear and distinguishable manner, at the latest at the time of the first interaction. Where other laws impose disclosure duties — including the UAE Personal Data Protection Law (Federal Decree-Law No. 45 of 2021) and the GDPR for EU data subjects — this notice operates alongside our Privacy Policy and Terms of Service.

Questions about this notice or about our use of AI:

Nexus Generated FZ-LLC RAKEZ License No. 5034603 VUET1700, Compass Building - Al Hulaila, Al Hulaila Industrial Zone-FZ, Ras Al Khaimah, United Arab Emirates Email: legal@odiseo.ai

We may update this notice as the platform and applicable law evolve. The "Last updated" date at the top reflects the current version. The English version of this notice is the binding version.