Is there human oversight of AI decisions?
Yes. amaise is designed as a human-in-the-loop system. No autonomous decisions are made by the AI.
Level of automation: Information preparation and recommendation. The AI presents classified documents, extracted data fields, and summaries to human claims adjusters.
Review and correction: Users can view, modify, or reject any AI-generated result.
Optional review stages: At the customer’s request, manual review or spot-check stages can be integrated into the pipeline to specifically verify AI results.
No automatic approval: No AI result triggers an automated payment, rejection, or legal action without human approval.
This design complies with GDPR Art. 22 (the right not to be subject to a decision based solely on automated processing) and anticipates the EU AI Act’s requirements for human oversight of high-risk AI systems.