Member Article

As employers begin to embrace Artificial Intelligence (AI) within the workplace, its growing use within the recruitment process means that the potential risks of its implementation need to be considered.

The Responsible Technology Adoption Unit (RTA) has published guidance to assist HR and recruitment organisations on how to use AI responsibly. The guidance encourages businesses to consider:

• the purpose of implementing AI at any stage in their hiring process;
• conducting an impact assessment of proposed AI systems which can evaluate the wider effects that a system might have on equality, data protection and human rights;
• implementing assurance mechanisms such as performance testing and bias auditing.

The RTA guidance urges employers to train employees who will be operating the software to better understand it and provide them with clarity around the recommendations it provides.

Employers could look to use AI for the following purposes:

• Sourcing candidates from multiple platforms and posting targeted adverts to reach the right audience;
• CV filtering to identify key skills, qualifications, and phrases relevant to the specific role;
• Initial candidate assessments such as automated video interviews.

However, whilst, technology can certainly assist employers, we are a long way from a fully automated recruitment process. Given the potential risks of relying on AI, human involvement is still key. 

Bias and risk of discrimination

Employers should, for instance, be mindful of the potential risks of using AI within their recruitment plan. Removing the human element can for example, decrease the subjectivity of judging candidates but it does not prevent biased algorithms from impacting the decision-making. This is because AI systems are often trained using historical data of what might have been considered successful for that outcome in the past but potentially not in a more progressive present. For instance, an algorithm that is trained with data sets derived from trends in CVs of current employees will seek out similar applicants. If the current staff is mainly male, then this is likely to result in the algorithm preferring men and unintentionally discriminating on the grounds of sex.

When using AI, employers must ensure they are still complying with the Equality Act 2010 which creates a legal obligation to ensure that applicants with disabilities, conditions and impairments are not disadvantaged throughout the hiring process. This includes reviewing what limitations the systems may possess and how these might apply to candidates including whether reasonable adjustments need to be made. For example, if the technology is screening written or video responses, consider what additional steps might be required to ensure that disabled candidates are not disadvantaged. Software that is screening applications could inadvertently exclude someone experiencing issues arising from Tourette’s, dyslexia or dyspraxia or heightened anxiety. Employers need to ensure that if the software cannot adjust its response to accommodate disabilities, a human can step in. Otherwise, they might face the risk of indirect discrimination, as well as a lack of reasonable adjustments on their part.

Data Protection

Finally, it is key for employers to be transparent about the use of AI systems when recruiting to enable applicants to contest machine made decisions. If candidates are not aware that a system is being used, they cannot possibly know whether to challenge outcomes. In any event, the UK GDPR restricts employers from making solely automated decisions that have a significant impact on job applicants and workers except in limited circumstances, such as where the decision is necessary for entering into or performing the employment contract or where the data subject has consented. Employers are unlikely to meet these exemptions and should therefore always ensure that there is human influence on the outcome in any employment decisions involving technology.

Conclusion

In conclusion, the integration of AI in recruitment processes looks to enhance efficiency and innovation. However, it is imperative that employers navigate this terrain with a conscientious approach that prioritises fairness, transparency, and compliance with legal standards. The RTA’s guidance serves as a beacon, illuminating the path towards responsible AI adoption that respects human rights, data protection, and equality. 

Employers must remain vigilant, ensuring that AI tools are used to augment human decision-making, not replace it. By maintaining a balance between technological advancements and human oversight, organisations can harness the power of AI while safeguarding against biases and discrimination, ultimately fostering a diverse and inclusive workforce. 

 

Link to article