AI Chatbots Struggle to Maintain Reliability in Extended Conversations

A comprehensive new study analyzing over 200,000 AI conversations has revealed a troubling pattern that could reshape how we think about AI reliability. Research conducted jointly by Microsoft and Salesforce shows that AI chatbots experience a dramatic 112% drop in reliability during extended conversations, despite performing exceptionally well on single-turn interactions. According to reports from this groundbreaking research, while AI models excel at answering isolated questions, they struggle significantly to maintain accuracy and coherence as conversations progress through multiple exchanges.