GPT-4 Summarization: Claims Processing Concerns

What’s your experience with using OpenAI’s GPT-4 API for document summarization? I just saw New AI Travelers using it to fast-track claims processing, and honestly, it’s a bit unsettling. Their internal estimates show a 60% reduction in processing time for simple auto claims—that’s a huge shift in operational efficiency, but I’m also wondering about the potential for biases creeping into those automated decisions.