The road to smart recruitment: Ethical implications of AI and automation in hiring

This is branded content.

Recruitment can often be challenging for even the most highly skilled teams. In recent years, improvements in communications technologies, and an increasingly interconnected world, have led to twin problems facing recruitment teams - dealing with an ever-increasing volume of applications, and an increasingly complex hiring landscape.

Innovations in digital technologies have empowered HR recruitment professionals, from recent graduates of a Masters in HR to seasoned recruitment professionals, with the tools necessary to extract and synthesise complex insights from large volumes of recruitment data. In a world where not all applicants have the same level of technical capability, recruiters must scrutinise the roles that AI and machine learning have in the recruitment process.

What are the ethical implications of using AI to streamline and optimise the hiring process? What are the limitations and biases that are present in these tools?

How can we best balance the risks of AI with the significant efficiency benefits it presents to teams?

Let's discover how AI is changing the face of the recruitment industry, and what stakeholders, from potential employees to senior management, should be aware of when using these powerful tools.

Modern recruitment uses several digital tools to enhance the workflow of HR professionals, using a range of OCR (optical character recognition) and parsing tools to rapidly extract and summarise relevant information from candidate CVs, resumes, and letters. In a world where employers may receive dozens or hundreds of applications from across the globe, being able to parse and streamline this information to make data-driven decisions is crucial.

AI and machine learning are not just embedded in automated document reading, however. Tools such as candidate matching can use algorithms that consider a range of factors to identify the potential fit that a candidate may have in their role. Algorithms can also be used as a form of digital peer review, assessing recruitment decisions and evaluating for the potential unconscious bias that may appear in human-based recruitment.

At an operational level, recruitment enhanced by AI and ML tools can yield significant efficiency benefits, reducing the time it takes for HR recruiters to manually extract information from hours to minutes.

AI recruitment tools can also be incredibly powerful in tackling unconscious bias that may exist within candidate pools. People are human, and it's natural for some level of unconscious bias to exist within recruitment processes. Using AI tools, at least some level of this bias can be mitigated by algorithms, allowing for a recruitment process to consider data, rather than demographics.

Many companies, such as global consumer goods producer Unilever, have leveraged the power of AI to great success within their recruitment processes. In one case, Unilever's team had to manage an applicant pool of 250,000 university graduates, to select just 800 for a leadership program. Their existing process, consisting of paper and manual processes, could potentially take four to six months to select candidates - not only massively time-consuming but highly inefficient for recruitment teams, particularly as these graduates were in high demand by a range of companies.

The solution? Apply an AI tool that utilises video interviews and assessment technology to create data-led outcomes. This was highly effective at creating a process that not only supported potential candidates but also improved Unilever's average candidate onboarding time from months to a matter of weeks. Much can also be said of the cost-benefit of the tool used, with Unilever reporting more than one million pounds ($1.9 million AUD) in cost savings.

While it's clear that artificial intelligence and machine learning tools can help drive smart recruitment decisions, it's important to recognise that algorithms, like humans, are not faultless. In recent years, there have been many incidents that have highlighted the challenges that job seekers face when applying for work through AI-augmented processes.

A common saying known simply as 'you are what you eat' has origins going back nearly two centuries, with its origin often being attributed to French polymath Jean-Anthelme Brillat-Savarin. While Brillat-Savarin had meant his words in terms of gastronomical application, rather curiously, it can also be applied to the way that contemporary AI models are constructed and implemented.

As many early models have shown, providing training data to AI models that contain biases will inherently create bias in a model. For example, facial recognition technology has been plagued globally by a poor capability to recognise those with coloured or dark skin tones - leading to the risk of racial profiling by those that use these tools.

While there are risks in the underlying data, in recent times, a new avenue has emerged in recruitment bias - the ability for savvy applicants to game generative AI systems to land positions at companies, such as encoding instructions to a model, hidden in font that is invisible to human recruiters, but identifiable by OCR tools.

A lesser-seen risk is the challenge of managing data generated by smart recruiting tools. While it's handy that data can be extracted by AI tools, it's important to consider the risks that are present to applicants and future employees if a recruitment platform is compromised by malicious actors.

Ultimately, risks are present in all forms of recruiting. It's critical that businesses not consider smart recruiting tools as a way to eliminate recruiting teams in their entirety, instead, envisaging them as a tool that they can use to augment the recruitment process.

Some employers are balancing the efficiencies of AI with the complex capabilities of human recruiters by establishing a workflow that takes the best parts of both roles and combines them to create a highly efficient team. For example, an employer may empower their teams with clear guidelines on how AI can be used in recruiting, promoting the ethical use and handling of data, right through the customer lifecycle.

Smart recruitment tools and human recruiters can be incredibly potent tools in recruiting and retaining the best possible staff for teams. While there are indeed challenges when using AI tools, they present significant opportunities for HR recruiters to transform the way that they recruit, for the better.