Catching Mistakes By AI Language Models Just Got Easier

This new research tool claims to empower users with transparency, helping content consumers trust AI-generated information.

On August 5, 2025, Microsoft Research introduced a new framework called “VeriTrail.” What makes this notable is the claim Microsoft has made about this new tool - that it’s designed to detect when artificial intelligence (AI) language models (LM) make things up (called “hallucinations”) and to trace where each piece of information came from in multi-ste…