Putting D&I at the heart of AI
Ingrid Waterfield, August 23, 2017
AI could perpetuate the biases HR has been seeking to remove from systems, processes and data
“The revolution is upon us!” we are told by various commentators, and the shout is becoming louder, more frequent and more urgent as time goes by. Furthermore, you need to embrace the glorious revolution before it is too late and you’re left behind.
AI is coming, is already here (depending on your budget or your organisation’s desire for innovation), and is here to stay. It will transform the way we organise ourselves collectively and behave individually in practically every sphere of our lives.
Business is no exception. There is evidence to suggest that AI will save costs, drive up productivity, free us up for innovation, and fuel growth by providing products and services to a more diverse customer base.
However, HR professionals must also recognise that AI is only as good as the software written for it, and the data sets that it examines – it essentially looks at the past to develop behaviours to predict the future. But what if the software and/or data are inherently biased? What if there is subtle and insidious confirmation bias 'baked in'? AI could perpetuate the very biases that HR has been seeking to tease out of its systems, processes and data.
But just as biases can be programmed in, they can also be programmed out. AI could be a fantastic tool to augment HR’s functionality and help drive diversity and inclusion strategies. These key steps will help HR put D&I at the heart of an organisation’s AI/business transformation strategy.
1. Understand how digital labour will affect your business
How will digital labour affect the diversity of your existing workforce and the ones you want to attract and retain? Be creative about how data can enhance the ability to attract, develop and engage diverse talent and encourage inclusive behaviour.
Will you need to recruit a different type of person? How will you reach and attract them? This is especially important for global multinational organisations. For example, AI could analyse the language used in existing job descriptions and the insights gained could help maximise their appeal across different demographic groups that you want to draw to your organisation.
2. Use AI to better understand your workforce and provide an employee experience that suits their specific needs
Use AI to estimate an individual’s likelihood of staying. One of our clients had an issue with staff attrition of up to 40% in some areas so they used data points to predict who was most likely to leave based on behaviour patterns changing over time. By identifying those most at risk of leaving they could act in advance and potentially avoid it happening. This approach had a significant impact on their bottom line.
Conversely, they were also able to adapt the data and understand more about what leads to an individual being a star performer and amend their recruitment assessment process accordingly.
3. Ensure you are a key stakeholder of your organisation’s AI/digital labour strategy
Incorporating these technologies into your business is a strategy decision, so it’s important that HR engages with colleagues across the company to:
- Define biases already inherent in existing activities that digital labour is replacing (e.g. judgement- and skills-oriented)
- Ensure there is consideration (ideally a specific work stream) focusing on adjusting any new digital activities to be free from bias
- Use AI to encourage inclusivity that creates innovation and creativity
- Replicate how a more diverse workforce provides a better customer journey.
D&I should not be a bolt-on to existing HR strategies, but at the centre of any business use of AI, because the benefits will be quickly realised by organisations looking to achieve a competitive edge.
Ingrid Waterfield is director of people and change at KPMG