23/04/2026
There may be “no ‘I’ in team” but there is certainly “AI” in bias. And also, unfortunately, bias in much of AI.
AI is increasingly being used within business to speed up processes and cut costs. It comes with a number of downsides too, one of which is bias.
It turns out that these large language models (LLMs), one of the main technologies behind AI as we know it, are trained on a vast amount of historical data. Unfortunately, it is data that has all kinds of human bias baked into it.
A classic example could be in CV sorting where the data an AI was trained on showed a preference for male candidates over female. Now it applies this same bias automatically. Not only would this expose you to legal risk under the Equality Act, but it could also mean you miss out on the best candidates due to the bias. We discuss this further in April's edition of
We also discuss keeping staff records, political tension in the workplace and more.
To read these articles in full, and find out the latest from the world of HR, check out our newsletter here:
You can’t spell bias without “AI” There may be “no ‘I’ in team” but there is certainly “AI” in bias. And also, unfortunately, bias in much of AI. AI is…
