Agaton - Policies

---

AI Policy

Last Update: 2025/01/23

Table of Contents

1. Introduction
2. AI Systems Overview
3. Data Processing for AI
4. AI Governance and Oversight
5. Transparency and Explainability
6. User Rights and Controls
7. AI Security Measures
8. Ethical AI Use
9. Compliance with Regulations
10. Incident Response
11. Updates to this Policy

1. Introduction

This AI Policy outlines Agaton Technologies AB’s (“Agaton,” “we,” “us,” “our”) approach to the development, deployment, and use of artificial intelligence (AI) systems within our services. This policy should be read in conjunction with our Terms of Service, Data Security Policy, and Privacy Policy. In case of any conflict, the order of precedence shall be: (1) the Service Agreement, (2) the Data Processing Agreement, (3) the Terms of Service, and (4) this AI Policy.

2. AI Systems Overview

As detailed in Section 3.6 of our Terms of Service, our AI systems include:

• The Agaton Assistant, powered by advanced language models
• Proprietary Automatic Speech Recognition (ASR) technology
• Linguistic analysis systems

3. Data Processing for AI

In alignment with Sections 3.6 and 4 of our Terms of Service:

• We process user data to train and improve our AI systems, subject to explicit user consent
• All data used for AI training is anonymized and aggregated
• We adhere to data minimization principles, collecting only what is necessary for service provision and improvement
• Data processing for AI adheres to GDPR requirements and the EU AI Act

4. AI Governance and Oversight

As per Section 3.6 of our Terms of Service:

• We maintain appropriate human oversight of all AI systems
• A dedicated AI Ethics Committee reviews all major AI developments and deployments
• Regular audits are conducted to ensure AI systems operate as intended and within ethical boundaries
• We have established clear lines of responsibility and accountability for AI-related decisions

5. Transparency and Explainability

To ensure transparency:

• We clearly communicate to users when they are interacting with AI systems
• We provide explanations of how our AI systems make decisions, to the extent possible without compromising system integrity
• Regular reports on AI performance and impact are made available to stakeholders

6. User Rights and Controls

In accordance with Section 5.8 of our Terms of Service:

• Users have the right to opt-out of having their data used for AI training and improvement
• Users can request human review of significant decisions made by AI systems
• We provide tools for users to understand and control how AI systems use their data

7. AI Security Measures

Aligning with Section 3.4 of our Terms of Service:

• AI systems are subject to the same rigorous security measures as all our other systems
• We implement specific safeguards to prevent AI-specific vulnerabilities and attacks
• Regular security audits are conducted on our AI systems

8. Ethical AI Use

We are committed to the ethical use of AI:

• Our AI systems are designed to avoid bias and discrimination
• We regularly assess the societal and environmental impact of our AI systems
• We have established ethical guidelines for AI development and use, which all employees must follow

9. Compliance with Regulations

As stated in Section 3.6 of our Terms of Service:

• We comply with the EU AI Act and other applicable AI regulations
• Our AI systems are designed with privacy-by-design and security-by-design principles
• We maintain documentation of our AI systems as required by regulations

10. Incident Response

In line with our Data Security Policy:

• We have specific procedures for handling AI-related incidents
• Any significant AI system changes or incidents are promptly communicated to users
• We cooperate fully with authorities in case of any AI-related investigations

11. Updates to this Policy

As with our other policies:

• This AI Policy will be reviewed and updated regularly to reflect changes in our AI systems and applicable regulations
• Users will be notified of any significant changes to this policy
• The latest version of this policy will always be available on our website

By using our services, you acknowledge that you have read and understood this AI Policy. If you have any questions or concerns, please contact our Data Protection Officer at dpo@withastrid.com.

This AI Policy is effective as of 2025/01/23 and will remain in effect except with respect to any changes in its provisions in the future, which will be in effect immediately after being posted on this page.