Since the arrival of ChatGPT, heralded by some as the dawn of the age of AI, the subject of artificial intelligence has provoked strong reactions. For some, it will make life better across the board: boosting productivity, freeing human beings to employ more empathy and creativity, and helping us to solve existential problems such as the climate emergency. For others, it is the existential threat, ushering in mass job losses or even bringing an end to the human story and the start of the robot one.
What we can all agree is that AI will revolutionise our work and our lives. For HR professionals concerned with inclusivity – and a whopping 98 per cent agree that EDI is important to business strategy, according to a poll commissioned on Monica Motivates’ behalf by OnePoll – the question is: will AI help or hinder us in our efforts to create more inclusive workplaces?
AI has a bias problem. On numerous occasions, facial recognition technology has misidentified people – principally women and people of colour – sometimes with disastrous consequences, as when Robert Williams, a Black man, was arrested in Detroit without having done anything wrong.
This is a consequence of lack of diversity, as research has shown: without the input of historically underrepresented groups in the creation and development of AI, including in the datasets that train it, AI risks ending up suffering from the same biases its creators have. Though OpenAI has worked hard to put in place guardrails to eliminate bias from AI, they can be circumvented.
If we introduce AI into the workplace too quickly or too carelessly, therefore, we may risk entrenching existing biases, with negative consequences for inclusivity. AI could intensify bias as regards hiring: Amazon scrapped its recruiting tool when it showed bias against women. And researchers at Cambridge have suggested that AI could ultimately increase uniformity rather than diversity in the workforce, excluding those who do not resemble the majority.
The key word in the above is ‘could’. Because AI also has the potential to take more menial, more manual, tasks out of the hands of HR professionals, liberating them to call on their defining characteristics: their empathy, creativity and deep concern for the individuals in their workplace. They could do their jobs more effectively as a result and that has beneficial implications for inclusivity and more.
And though AI is not itself free of bias, it could still be used to flag human biases, especially in recruitment, if used carefully. It could improve the inclusivity of the language used in job postings, support employee engagement and boost diversity, which is closely intertwined with inclusivity. AI could, for example, be trained to focus on skills, qualifications and experiences rather than factors that might be influenced by biases, such as an applicant’s name, age or appearance. It could be used to gather and crunch pay data and identify inequities – something of huge value in particular for large organisations, where such a task would be laborious (and prone to error) for a human.
AI could also improve communication, by providing instant feedback on the tone, style and linguistic inclusivity in employees’ writing. It could identify patterns in levels of engagement, gathering data, analysing it and indicating where change is needed on an ongoing basis. And it could support employees working in their non-native language and neurodivergent employees who may have barriers to communication. This increased ability to communicate effectively could greatly increase individuals’ sense of belonging.
AI is a powerful tool. That much is clear. Whether it helps or hinders HR professionals depends on how they use it and whether its shortcomings are fully understood. In the right hands, it could have an overwhelmingly positive impact on workplace inclusivity, promoting a more nuanced and complete understanding of the workforce and supporting effective interventions while empowering individuals with disadvantages in communication.
The internet is a useful parallel, and not just because of its game-changing impact. The internet can increase polarisation, bring people into contact with unhealthy content and, in short, do harm. But it has also brought education to the fingertips of millions, brought down borders, greatly increased accessibility, connected individuals round the world and so much more besides. It’s not the ‘what’ but the ‘how’ that counts most.
So it’s my firm belief that AI can benefit HR professionals as they strive to make the modern workplace inclusive for everyone. Biases can be weakened, guardrails can be strengthened and bugs in the system can be discovered and patched. And in the meantime, we can and must make sure that we use AI with our eyes open, that we accept its shortcomings and weaknesses and that we drive EDI so that future iterations of AI become increasingly humane.
HR professionals share this view. In the survey we commissioned from OnePoll, when asked for their views on AI, just 7 per cent said they perceived it as a threat, suggesting that the vast bulk of those working in HR see AI as having a positive role in the workplace. This is encouraging: a positive attitude is a precondition for taking an active role in making the technology something that will serve us all.
Monica McCoy is CEO and co-founder of global consultancy Monica Motivates