Many people are increasingly using generative AI to help with tasks at work. While doing so is certainly convenient and can have dramatic improvements to productivity in roles like software engineering, it also raises some potentially serious issues, one of which is around accountability.
Who is accountable for decisions made and actions taken where AI is used to provide advice, write reports, make recommendations and the like? There has been a slew of LinkedIn posts recently about the lack of accountability at the leadership level of organisations, and the increased use of AI in executive-level decision-making makes this an even more critical and timely issue.
I’ve spent the past couple of years presenting around the world on how I think AI will be used in hearing healthcare services, focusing on how AI will enhance the capabilities of audiologists rather than replace them. The analogy I have used to help audiologists better understand AI’s role in the clinic is to think of AI as a clinical intern who has a lot of knowledge but is prone to making mistakes and misunderstandings. An intern won’t replace the need for an experienced audiologist’s expertise and responsibilities, nor will the audiologist give critical decision-making responsibility to an intern, but the intern can offload work and provide their own perspective on issues that the clinical audiologist can choose to either act on or ignore.
The key is that the audiologist, and not the intern, is ultimately responsible and accountable for the care of their client/patient. Any work or recommendation that the intern makes must be reviewed and approved by the more experienced audiologist. This is the same with AI in the workplace: any AI system should be treated with no more authority than an intern, where the AI’s work must be reviewed for mistakes, misunderstandings and ill-formed recommendations so that they can be corrected before any damage is done.
Which brings us back to accountability. No leader would ever abdicate accountability to an intern, and no one should do that with AI either. To do so would be to undermine one’s credibility within the organisation, which according to recent reports is already a concern for many executive teams in large organizations.
This can be tricky for people to intuit because the benefit of using of AI at work can be profound and powerful, but no leader should have AI provide critical analysis or advice if that leader can’t identify mistakes made or nuances not understood. To do so would be to abdicate responsibility and accountability.
The quote in the image above, from the book The AI-Driven Leader, describes one of the risks of using AI as a leader and it summarises this whole discussion nicely. The book’s author Geoff Woods describes the role of the user of AI as the Thought Leader and the role of AI as the Thought Partner. This is a nice way of thinking about the role that AI plays in decision-making and critical thinking, as someone to test ideas on or hear suggestions from, and it is not too different from my approach to thinking of yourself as the Accountable Leader and AI as the Knowledgable Intern.
So, next time you use AI for a critical task, think of it as a really smart intern rather than an expert that can overrule or replace your own decision-making.