By now, most of us are aware of the pervasive nature of implicit human bias. Organizations worldwide have invested significant resources in bias training initiatives, despite hundreds of studies dating back to the 1930s1 that consistently demonstrate their ineffectiveness in reducing bias, modifying behavior, or mitigating its impact in the workplace.
Similarly, the acknowledgment of AI bias and its potential adverse effects on underrepresented groups is becoming increasingly prevalent. As someone deeply engaged in addressing these issues, I’ve developed an AI-powered tool aimed at fostering inclusivity. At first glance, this might seem paradoxical. One might ask, ‘Isn’t AI inherently biased?’
The truth is, AI inherits biases from its human creators. Whether it’s candidate shortlisting algorithms or generative chatbots like ChatGPT, the decisions AI makes are shaped by the criteria and datasets provided by humans. However, this also means that with careful curation and training on objective, validated, and non-biased inputs, AI can transcend bias and promote inclusion.
As a behavioral scientist, I’m fascinated by the factors influencing human decision-making. A recent study2 I came across found that participants in experiments, when given biased suggestions by a fake AI, continued to exhibit the same bias in their decision-making even after the AI guidance was removed. The findings from that study highlight the profound impact AI interactions can have on long-term human behavior. With this being true, it then stands to reason that non-biased AI models can influence non-biased human decision-making processes too.
This leads us to an intriguing proposition: if traditional bias training fails to produce meaningful change, but we can build AI that makes unbiased decisions, and AI can influence human behavior positively, then we have a powerful tool at our disposal for driving inclusive practices and equitable outcomes.
The convergence of AI, bias awareness, and behavioral science presents a compelling opportunity for HR professionals to champion diversity, equity, and inclusion in the workplace. By leveraging non-biased AI models, we can reshape decision-making processes and create a world where everyone is included and can thrive.
Dr. Liz Wilson is a behavioral scientist, organizational transformation expert, and founder of Include Inc. Originally from Australia and now based in the United States, Dr. Liz is well known for her authentic, honest, and pragmatic approach to everything she does. This includes her simple, yet powerful, Include Change Method that has achieved amazing results for her clients over the span of her 25+ year career. Dr. Liz has transformed the ways of working and cultures of dozens of global organizations including; major airlines, banks, consumer goods, telecommunications, mining, rail, tech and medtech companies.
1Dobbin, Frank, and Alexandra Kalev. “Why doesn’t diversity training work? The challenge for industry and academia.” Anthropology Now 10.2 (2018): 48-55.
2Vicente, L., Matute, H. Humans inherit artificial intelligence biases. Sci Rep 13, 15737 (2023). https://doi.org/10.1038/s41598-023-42384-8
