Equality in AI: The EHRC’s Updated Approach to Regulating AI in the Workplace

As the potential for AI technologies to transform the workplace comes into ever greater focus, and as the capabilities of those technologies continues its rapid advance, concerns over the potential threat to workers’ rights and the challenges of regulating the application of AI technologies have increased. 

In an effort to address those concerns, the Equality and Human Rights Commission (EHRC) has recently published an update on its regulatory stance towards AI, outlining its approach to ensure protection of equality and human rights in the UK. 

As an independent regulator tasked with promoting diversity and inclusion in the UK, the EHRC holds a statutory mandate to advise the government and Parliament on equality and human rights issues and has a regulatory remit that spans all sectors. 

The EHRC’s update follows the government’s White Paper on AI regulation (the “White Paper”), published March 2023, which underscored the need for collaboration between government, regulators and businesses to oversee AI implementation. The White Paper envisioned existing regulators covering specific sectors to regulate AI in the context within which it is used. 

EHRC’s approach to regulating AI to date

Addressing the equality and human rights impact of AI has been a strategic priority for the EHRC since 2022, focussing on key issues such as breaches of the Equality Act 2010 and Human Rights Act 1998. 

For instance, most recently the EHRC funded the case of Uber Eats driver Pa Edrissa Manjang as part of their powers to provide legal assistance to victims of discrimination under section 28 of the Equality Act 2006. Mr Manjang alleged that the facial recognition checks required to access his work app were racially discriminatory and resulted in an inability to secure work due to the continuous difficulties he faced when attempting to complete verification checks. He received a financial settlement as a result of his claim.

In addition, the EHRC, alongside the CMA, Ofcom, Information Commissioner’s Office (ICO) and FCA, has participated in the Digital Regulators Cooperation Forum (DRCF) since 2020, a forum for effective collaboration on digital regulatory matters.

The EHRC’s updated approach

The EHRC’s updated approach to AI regulation incorporates the White Paper’s proposed five non-statutory principles: 

• safety, security and robustness
• appropriate transparency and explainability 
• fairness 
• accountability and governance 
• contestability and redress

While the EHRC expressed its support for this principle-led approach, it also criticised the omission of equality and human rights considerations and the lack of additional resource allocated to meet the evolving and more onerous regulatory demands expected by the government. Even so, the EHRC identified a potentially enhanced role for itself working with and across all regulators in order to ensure that equality and human rights considerations are consistently applied in all areas in which AI is used. 

In particular, the EHRC identified a leadership role in relation to the fairness principle, noting a strong alignment between their regulatory remit and the ambitions of the government. Namely, the White Paper notes the expectation for regulators’ interpretation of ‘fairness’ to consider relevant law and regulation, such as whether AI systems produce discriminatory outcomes which contravene the Equality Act 2010 and/or the Human Rights Act 1998.

Looking ahead, the EHRC has no plans for the remainder of its strategic plan to develop dedicated guidance around the White Paper principles due to resource constraints. Instead, the regulator plans to focus on, among other things, the use of AI in recruitment practices and developing solutions to address bias discrimination in AI systems. 

Recruitment is designated a ‘high risk’ activity under the EU AI Act, so it is perhaps unsurprising that this is an area of specific focus for the EHRC. It will also be interesting to see whether the EU's approach may drive up standards in this area.

Comment

The EHRC’s update is a welcome (if somewhat limited) commitment to navigating the challenges of AI regulation while ensuring protection of equality and human rights in the workplace. 

It remains to be seen whether the EHRC will - within its current size and resource constraints - be able to meet the challenge put down by the government in its call for a flexible and adaptive regulatory approach, and whether it can keep pace with the dizzying pace of developments in this area. And as the technology becomes more sophisticated, so will increase the complexity of the enforcement challenge.

There will inevitably be lags between advancement/implementation of new AI technologies and subsequent regulatory intervention – but the extent to which we see regulators like the EHRC playing ‘catch up’ in this area will depend in part on the level of financial and human resource investment that this and any future government is willing to make.