Ethical AI Use Policy

1. Purpose and Scope

This policy sets the framework for how we develop, deploy, and govern AI responsibly. It applies to all employees, contractors, and partners working with Unsupervised's AI systems, particularly those analyzing or processing customer data.

This policy is designed to ensure:

• Responsible data handling
• Protection of user privacy
• Transparency in design and deployment
• Fairness and non-discrimination
• Accountability through oversight and governance

2. Responsible Data Handling

We recognize that data is entrusted to us, not owned by us. Our mandate is to connect, store, and process it with integrity.

Our commitments:

• Collection: Only connect to data that is necessary and relevant to the end user's goals, with explicit customer consent.
• Storage & Retention: Protect data with encryption and access controls, retaining it only as long as legally or contractually required.
• Security: Maintain strong authentication, authorization, and monitoring practices, with regular risk assessments and remediation.

3. Privacy and User Control

Privacy is fundamental to ethical AI. We embed privacy-by-design into our systems and empower customers to control their data.

Our commitments:

• Comply with privacy regulations (e.g., GDPR, CCPA).
• Honor flags for PII and Protected Data and give customers the option to not include that information in analyses.
• Provide customers with clear ways to request corrections, deletions, or restrictions on processing.
• Require third parties who handle data on our behalf to meet the same privacy standards.

4. Transparency and Explainability

AI should never be a "black box." We communicate openly about how our systems work and what their limitations are.

Our commitments:

• Clearly document and share the purpose and functionality of AI agents and AI analysis.
• Distinguish between automated outputs and those reviewed or adjusted by people.
• Inform customers of model limitations, edge cases, and potential risks.
• Strive to make AI outputs understandable so customers can use them with confidence.

5. Fairness and Non-Discrimination

We aim for equitable outcomes and actively work to detect and reduce bias in AI.

Our commitments:

• Regularly evaluate models for harmful bias and take corrective action.
• Collaborate with customers to ensure input data is representative and inclusive.
• Conduct periodic fairness audits and adjust models as needed.

6. Accountability and Governance

Ethical use requires oversight, responsibility, and a culture of integrity.

Our commitments:

• An internal Ethics & Compliance Team (ECT) reviews AI projects prior to deployment.
• Employees are trained in data handling, privacy, and AI ethics, with secure channels to report concerns.
• We maintain an incident response process for AI-related complaints, investigating and resolving issues quickly.

7. Continuous Review and Improvement

Technology and regulations evolve, so must our practices.

Our commitments:

• Review this policy annually and whenever major changes that we are aware of occur in law, technology, or best practice.
• Incorporate feedback from employees, customers, and partners into updates.
• Document and communicate all policy changes transparently.

8. Compliance and Enforcement

This policy is binding. Every team member and partner must adhere to it.

• Violations may result in corrective action, including termination of employment or partnership.
• This policy complements — but does not replace — applicable laws and regulations. Where conflicts exist, governing laws take precedence.

9. Contact and Reporting

For questions, guidance, or to report concerns:

• Ethics & Compliance Team: ect@unsupervised.com
• Privacy Office: dataprivacy@unsupervised.com