ChatGPT in HR: De-biasing your recruiting process & creating AI policy

Bias in a recruiting process is not only detrimental to an organization’s diversity efforts; it also has business implications. According to McKinsey, companies fail to choose the candidate with the right talent for management roles 82% of the time. Why? Because systemic, unintentional and pervasive biases mean the employer team is relying on opinions, rather than fact.

There are a number of core principles for de-biasing a recruiting process – and AI has the potential to be disruptive to that process in both positive and negative ways. When de-biasing the recruiting process, we begin by walking through from beginning to end, and understanding where decisions are being made and how biases could shape those decisions.

Organizations integrating AI into their recruiting process should take AI prompts and outputs into consideration as part of that walk-through. It’s possible (and in many cases inevitable without careful prompt writing) for AI to introduce bias into a process, so thinking critically about how AI is being used and what outputs it’s generating is key. Organizations should consider DEI efforts as part of their training on how to write effective AI prompts, because AI does have DEI data integrated.