Artificial Intelligence Policy

This Policy describes how Ruum plans to use artificial intelligence (AI) in the Service, the safeguards it will apply and the rights of users. It complements the Privacy Policy and the Terms and Conditions of Use.

---

Table of contents

1. Current status
2. Planned AI features
3. Models and providers
4. Data processed with AI
5. Safeguards and limits
6. Human decisions and oversight
7. Your options and rights
8. Changes to this Policy

---

1. Current status

AI features are not yet generally integrated in production. This Policy is published in an early and transparent manner. Before activating an AI feature that processes your data, Ruum will update this Policy and, where appropriate, obtain your consent.

---

2. Planned AI features

Ruum plans to incorporate, among others, the following features:

• Automatic classification of chapters and line items.
• Extraction of information from documents or texts.
• Content generation (e.g. line-item descriptions).
• Conversational support assistant.
• Suggestions for preparing quotes.
• Workflow automations.
• Semantic search using vector representations (embeddings).

---

3. Models and providers

Initially, Ruum may use AI models from third-party providers (for example, OpenAI, Anthropic or Google), in addition to its own models. When external providers are used, Ruum will:

• Enter into a data processing agreement (DPA) with them, under which they act as processor.
• Require, to the extent the provider offers it, terms of non-use of your data to train their models and limited or zero retention of the data sent.

The list of active AI providers will appear in the Subprocessors Annex.

---

4. Data processed with AI

Ruum will apply the principle of minimisation: only the information necessary for the requested function will be sent to the models. Whenever technically possible, sending unnecessary personal or confidential data will be avoided, or anonymised or pseudonymised data will be used.

AI usage logs (prompts, responses and inference metadata) may be associated with the account or company for the operation, security and improvement of the Service.

---

5. Safeguards and limits

• AI is a support tool: its results may contain errors or inaccuracies and must be reviewed by a professional before use.
• Ruum does not guarantee the accuracy of AI-generated content.
• Ruum will not use your confidential content to provide advantages to third parties or for purposes other than those described.

---

6. Human decisions and oversight

AI features will not make decisions with legal or significant effects on you without the possibility of human intervention. The professional user retains control and ultimate responsibility for the quotes and documents.

---

7. Your options and rights

• When activated, you will be able to know which features use AI.
• When an AI feature is based on your consent, you will be able to not activate it or withdraw your consent.
• You retain the rights described in the Privacy Policy.

---

8. Changes to this Policy

Ruum will update this Policy as it activates AI features or providers change. The version in force is the one published on the site, with its update date.