Making AI implementation in HR more ethical
How (and Why) I created a canvas for ethical implementation of AI in HR
Why ethical AI is important, especially in HR
AI seems unstoppable these days. It’s everywhere. It dominates the discussions at the Davos Conference, with some of the most notable entrepreneurs and thinkers presenting to world leaders on the challenges, opportunities, and risks of implementing AI for (hopefully) humankind's benefit.
And one of the biggest challenges of AI is building trust. Which is, coincidentally, one of the biggest challenges of HR as a discipline. Therefore, it follows that the challenge would compound when trying to apply AI to HR.
But where to start? Where to put the focus? What perspectives to include?
There doesn’t seem to be a clear framework to do that, so I tried to contribute to the discussion by creating a tool HR leaders can use to plan their implementation of AI-related initiatives.
Enter the HR AI Ethics Canvas
Taking the Business Model Canvas as an inspiration, I tried to put together this tool that I am now releasing into the world, building in the open in hopes that other HR leaders can use it to build more ethical AI implementations.
This is version 1 of it, and it is a constantly evolving landscape, so I know it’s not going to be perfect, but I hope:
launching it now helps me gather feedback for improvements, and
using it inspires others to contribute their part to make AI more beneficial
My creation process
The creation process was, as is to be expected, heavily reliant on AI tools. Here’s my workflow, which resembles what I have been using for a lot of my recent writing:
Initial Research
First, I did some initial research for existing frameworks and tools using Perplexity.ai, which allowed me to identify many sources of inspiration, including:
Jasper’s AI Policy template: a simple and straightforward starting point.
AI Guardian’s AI Policy template: another great foundation to build upon.
OpenEthics’: a very thoughtful and exhaustive canvas for AI Ethics.
And an honorable mention for AI Ethicist’s growing collection of toolkits.
And of course, a lot of policies from companies that are pioneering their implementation of AI and felt it was their duty to share their findings with the world (Rolls Royce’s Aletheia framework is an example of that).
Brainstorming
Almost in parallel with the initial research, whenever I have an idea that’s a bit cloudier than I’d be comfortable with, I do a brainstorming session with Inflection’s Pi. This AI agent was designed to be uniquely personable and easy to interact with. It doesn’t just respond: it converses with you, analyzing your responses and asking follow-up questions that help you build clarity for yourself, so you can prompt it better.
Once the idea is a bit more polished, I can bring it to Claude and ChatGPT for further expansion of some of the concepts, and some content generation.
In particular, I used these tools to:
Identify a list of sections for the canvas that is exhaustive and reasonable. (Note: The list started quite long and redundant because different sources call the same things different names, or split concepts in different ways. So I did a mind mapping and whiteboarding exercise in Miro to help me narrow down the most representative ones, and determine a hierarchy of concepts.)
List the top 3 questions that you need to ask to complete all of these sections.
Come up with some illustrative examples to further explain each question.
And then, I went into Canva to create it, using a template I had created for my GPT Creator Canvas.
Next steps
Publishing it was the first step. Now that it’s out there in the open, I would love to gather as much feedback as possible from AI, Ethics, and HR professionals in the different communities I am part of, such as AIAgile, PeopleGPT, and many others.
Ultimately, want this to be a foundation that other people use to build better AI, and better HR. So feel free to share, borrow, and branch, I only ask you to give credit where credit is due.
And follow my journey on
as I continue to explore the intersection of HR, Technology, Product, and Agility to inform my practice.Speaking of practice, here comes the plug:
I recently launched my Go Fractional profile for Fractional Head of People roles, so any feedback, referrals, and sharing would be much appreciated!