While many of the risks associated with AI are now better understood, some organisations continue to experience undesirable outcomes with their integration.

Take this little doozy involving Microsoft 365’s Copilot, which resulted in widespread breach of users’ privacy protections. Which brings me to ask, ‘How are you assessing the risk associated with the implementation of your AI program?’

Or perhaps a better question is ‘Have you set yourself up for success?

In my Mastering Risk Workshop Facilitation training course, I recommend the following four actions to competently assess risks:

1. Ensure you have the right people in the room.
2. Establish key stakeholders and the risks they pose to AI initiative.
3. Identify external drivers of risk based on an environment scan.
4. Establish internal drivers of risk based on a capability assessment.

Assessing risks in an AI initiative, however, also requires a different approach from your standard strategic risk workshop, where you make sure you facilitate a fantastic risk conversation based on the insights you can draw out from all of the above and any existing data sets, like incidents, performance data or similar.  The approach I recommend for AI is a modified Hazard and Operability Study (HAZOP) technique, because the risks that can come with AI are often hidden from our minds.

The core HAZOP technique, explained here in greater detail, mentions modified HAZOPS as they relate to the chemical industry. However, I use the technique whenever I need to take the identification of risk way deeper than the norm. For example, during my recent facilitation of a range of workshops assessing a client’s digital transformation initiative I used the technique, including a situation-specific design of a set of guidewords to prompt people’s thinking.

It never ceases to amaze me how effective a good set of prompts alongside a visual representation of the initiative is for assessing risk. Time and again, the conversation starts with something like “Now that I think about it…”

Given the uncertainty that pervades AI initiatives for those early in their AI journey, techniques like a modified HAZOP approach will prove critical to your future success.

If you would like to learn more about how I help people navigate risk and AI, please reach out via info@bryanwhitefield.com.