Beyond the prompt: How high-performing auditors interact with AI
Register now to watch on demand
Hosted by Thomson Reuters
Ask AI the wrong way, and you get a handful of generic flags that go straight in the trash. Ask it the right way, and you get dozens of risk-ranked findings that hold up to scrutiny. Same model, same data — wildly different results.
This session unpacks why that gap exists, and introduces the CRAFT framework (Context, Role, Action, Format, Target), which is a practical way of thinking about what separates a reliable AI interaction from a generic one. Through examples, you'll see the same request produce dramatically different output depending on how it's structured.
You'll leave able to:
- Apply the CRAFT framework and what each element contributes to an AI interaction
- Understand why the same AI model can produce wildly different results depending on how it's used
- Recognize the difference between generic and well-structured AI output through real examples
- Evaluate AI output with a professional skeptic's eye — knowing what "good" looks like and what to question
No specific tool or platform required. The concepts apply whether you're using CoCounsel, ChatGPT, Copilot, or Claude.
Can't make it live? No problem. Register anyway and we'll send you the full recording to watch on your schedule.
* CPE credit is not offered.