In these times where every technology company is becoming an AI company, leaders need to be able to tell what’s real and what’s snake oil.

Here are 5 questions to guide your exploration of AI solutions, since they could sometimes be overstated by eager vendors.

Ask about

1. What’s going on algorithmically.
   • What is the nature of the “AI” involved? Can the models used be explained like to a 5 year old?
2. Interoperability.
   • How can this work with my existing infrastructure, products or my future roadmap?
   • What dependencies need to be factored in?
3. Explainability.
   • How are outputs verified?
   • When outputs say “What”, can you tell “Why”, “Why not” and “How”?
4. Production use cases.
   • Are solutions still experimental or are they functional products?
5. Return on investment.
   • Is the solution saving time, improving quality, or exposing new capabilities?
   • What’s the industry use case and addressable market for this capability, what’s your use case?