AI Transparency Notice

1. What this page is

This page explains how Mithilab uses artificial intelligence (AI) on this website, what AI systems are involved, and your rights — in line with the EU AI Act (Regulation (EU) 2024/1689) and the GDPR.

2. You are interacting with an AI

The chat assistant on this website is an AI system, not a human. We disclose this clearly because Article 50(1) of the EU AI Act requires it. Specifically:

• The chat widget is labeled "Mithilab AI" with an AI badge on every message.
• You can switch to a human at any time by clicking the "Human" button in the chat.
• If you ask the assistant whether it is human, it will truthfully tell you it is an AI.

3. Risk classification under the EU AI Act

We have assessed the AI system on this website against Annex I and the risk categories of the AI Act:

• Risk category: Limited Risk (transparency obligations only — Title IV)
• Applicable article: Article 50 — Transparency obligations for providers and deployers
• Intended purpose: Answer general questions about Mithilab services, suggest relevant solutions, and forward qualified leads to a human team member.

4. AI models we use

We do not train our own foundation model. We use Claude, a general-purpose AI model from Anthropic, accessed through their API.

• Model provider: Anthropic (PBC, USA)
• Model: Claude (Haiku / Sonnet)
• Hosting region: API request routed to Anthropic infrastructure (US). The website itself is hosted in the EU (AWS eu-west-1).
• International data transfer: Standard Contractual Clauses (SCCs) with Anthropic. Messages are not retained by Anthropic beyond the API request.

5. Your rights

When interacting with our AI, you have the following rights:

• Right to know — you are entitled to a clear explanation that you are talking to an AI (you have it on this page).
• Right to a human — at any point in the chat you may click "Human" and your conversation will be handed over to a person at Mithilab.
• Right to opt out — you can refuse to use the AI chat and contact us directly via email or the contact form.
• Right to delete — you can request deletion of your chat history at any time (see Privacy Policy).
• Right to complain — if you believe we are misusing AI, you can complain to the Dutch Data Protection Authority or, once operational, the EU AI Office.

6. No automated decisions affecting you

The AI assistant does not make any decision that has legal or significant effect on you. It cannot sign contracts, accept proposals, change account state, or process payments on your behalf. Every binding decision is made by a human at Mithilab on the basis of a written agreement.

7. Known limitations

AI systems are imperfect. You should be aware that:

• The assistant may produce incorrect or out-of-date information. Always verify pricing, timelines, and technical claims with a human before relying on them.
• The assistant only knows what is in its knowledge base (this website) plus general world knowledge. It does not know facts about you or your business unless you tell it.
• For sensitive or high-stakes decisions (legal, financial, medical, compliance), always speak with a qualified human professional.

8. What data the AI processes

Only the messages you type into the chat, plus a randomly generated session ID stored in your browser. We do not link your chat to your IP, email, or account unless you submit a contact form. Full details are in our Privacy Policy.

9. AI-generated content

Where blog posts, articles, or marketing copy on this website have been substantially drafted by AI, they are reviewed and edited by a human before publication. AI-generated images, if any, are labeled as such in the figure caption.

10. Contact and complaints

Questions, concerns, or requests related to our use of AI? Reach us at admin@mithilab.ai.