Factor’s 2026 GenAI in Legal Benchmarking Report found that only 22% of legal teams trust AI outputs. As AI-touched legal work becomes harder to avoid, the question is no longer simply, “Was AI used?” It is, “How was AI used?” What sources, context and constraints went into the tool? What role did the human play? And how was the result checked?
In this hands-on workshop, Factor’s Michelle Sherman explains why GenAI can produce fluent, convincing work even when its facts or reasoning are wrong. She shows how legal teams can review the process behind the answer, identify situations where hallucinations are more likely and verify AI-generated work without simply redoing the entire task.
Introduced by Factor's Global Head of Innovation Nicole Bradick and using practical exercises from Sensemaker Academy, the session introduces three ways to make verification more efficient: ask AI to show the exact source text behind its conclusion, assess how well each claim is supported and pressure-test the answer for faults or alternative interpretations. The goal is not to detect whether AI touched the work. It is to understand the answer and the human role behind it well enough to review the work appropriately.
• Why polished AI-generated legal work can still contain errors, hallucinations or unsupported claims
• How to review the inputs, process, output and human judgment behind AI-touched work
• Three practical techniques for verifying AI outputs more quickly and confidently
Find out more about Sensemaker Academy.