Responsible AI use and transparency
We follow EU GDPR and the principles of the upcoming EU AI Act. On this page, we explain how we design, train, and use AI in the Hello AI service safely and transparently.
Guiding principles
- Human oversight and control: users can override or stop automation.
- Technical safety: the system is tested, logged, and verified before deployment.
- Privacy and minimization: we collect only data necessary for operation.
- Transparency: we inform when AI is used in communication or automation.
- Non-discrimination: monitoring to detect and correct bias and distortions.
- Accountability and audit trail: events and changes are logged for later review.
How AI appears to users
- Hello AI notifies users when a message or call is AI-generated or processed.
- Automated decisions (e.g., call routing) are explained to users upon request.
- We aim to briefly explain what information the model bases its suggestions or decisions on.
What data AI uses
- Call and message metadata (timestamp, direction, technical status) for operation verification.
- Transcripts from calls may be temporarily processed for automated response or contact requests.
- Identity information (such as phone number) is needed to connect the right contact and prevent misuse.
- Model development primarily uses synthetic or anonymized data.
Risk management and monitoring
- Regular security audits and penetration testing.
- Continuous monitoring of AI behavior and outputs.
- Incident response procedures for security or bias issues.
- Regular updates and improvements based on feedback and new research.
Inquiries and complaints
Questions and complaints: support@sorwi.fi. Please specify whether your request concerns Hello AI or the website.