What’s HR’s role in AI governance?

The role for HR in AI

It’s well known that to gain value out of AI (or any other technology), we do need to consider how it’ll fit with the people. That includes:

• whether people have the skills to use it appropriately
• how well an AI-enabled solution supports the tasks that people do and their preferred ways of working (that elusive ‘culture’)
• how well the AI solution aligns with processes across the organisation.

Digitally mature organisations do these well and encourage their people to actively experiment and implement initiatives to respond to their customers’ changing demands. They encourage cross-functional collaboration and empower teams to act autonomously.

There’s also consensus that AI solutions should be underpinned by good AI principles. There are many good AI principles. For example, OECD’s AI principles call for AI that supports inclusive growth, fairness and privacy, transparency and explainability, robustness, and human accountability. Over 40 countries including the UK have already committed to these principles.

But when we talk about AI deployment or AI governance, it is rare to hear HR teams being actively involved in either. Through our conversations with HR practitioners, we know a few are members of cross-functional AI committees, proactively trying to influence their AI committees to take a more holistic approach, to think about the people and not just about the technology, or have been asked by their chief executives to think about how AI might change the shape of their workforce as well as hiring and skill development plans. In general, where AI is concerned, most HR teams tend to only think about effective AI use in HR, echoing CIPD’s findings from the digital transformation series. We found many examples of how HR teams leveraging technology effectively to deliver better HR services. But where it’s an organisation-wide transformation, this tends to be by a separate transformation programme management office (eg Innovation department) rather than HR.

Yet we know HR teams (large or small) do govern employment relationships. Not all HR teams have a reward or L&D specialist. But where HR exists, the responsibility for interpreting employment law and getting legal advice falls on HR. And so would the responsibility for implementing employment contracts, policies and processes that govern the employment relationships. When employment relationships sour, HR is responsible for dealing with disciplinary and grievance and reduce the risk of litigation.

Therefore, it’s not a stretch to say that HR has a role in AI governance wherever AI could affect the employment relationship. Governing employment relationships is already in HR’s remit. HR is simply using their existing governance skills in different ways. Such as:

1. Engaging in AI regulation discussion pragmatically and ensure ethics are meaningfully discussed.
2. Challenging AI governance by advocating for meaningful human oversight of AI.
3. Normalising good AI by embedding a culture where everyone organisation practises good AI principles like those promoted by OECD.

New resources for people professionals

As part of the CIPD’s work for the government-funded Innovate UK BridgeAI programme, we are developing resources to support HR in contributing to AI governance work in their organisations. This includes developing a guide in partnership with BSI on AI-related standards with illustrative examples for HR practitioners.

If you would like to suggest other AI governance resources that CIPD could signpost, then please contact Hayfa Mohdzaini (Senior Policy and Practice Adviser – Technology) at research@cipd.co.uk. We’re also interested in collecting a range of illustrative examples for non-technical readers explaining techniques for assessing the accuracy and trustworthiness of an AI-enabled solution.