The past year has brought some of society’s greatest inequalities to light, however it has also provided the opportunity for both individuals and organisations to reflect, reassess and address their processes towards Equality, Diversity and Inclusion. It is widely understood, not just due to perception but driven by statistical data that in the recruitment sphere discrimination has not evolved enough to reflect the modern world. Surveys in the last 12 months alone have highlighted that over a third of workers believe Covid-19 has delayed diversity efforts. It is universally accepted that more needs to be done, and could AI-assisted recruitment be part of the answer?
Are we biased?
One of the biggest challenges for implementing your Equality, Diversity and Inclusion strategy is unconscious human bias. As human beings we are hardwired to make intuitive decisions about other people and these unconscious judgements impact our attitude and behaviour towards other people, even without us noticing. Our brains use visual, verbal and behavioural clues, which often include age, gender, ethnicity or education to help us determine factors such as whether someone might be friendly or not. However in recruitment, this human survival technique actually hinders our ability to keep an open mind when hiring new talent. Could AI-assisted recruitment be the answer?
There are a number of different types of unconscious bias, triggered by social, memory and decision-making biases that impact us every day. Some of the most common ones in recruitment include:
- Confirmation bias – a tendency to favour or interpret new evidence as confirmation of existing beliefs or theories on a candidate, such as an interview confirming a CV judgement.
- Personal similarity bias – this common bias is where a hiring manager seeks to favour candidates who are similar to themselves, such as where they are from, educational experiences or background, not necessarily job related.
- Halo effect – this is a type of cognitive bias, in which the hiring manager is influenced by how they feel from the overall impression of a candidate, assuming their skills based on likeability.
Throughout the sourcing, sifting and interview recruitment stages, studies have shown that our unconscious bias is impacting the talent we take on and our efforts to grow a more diverse and inclusive talent pool. Whilst training, awareness and specific initiatives to pioneer the agenda can support us in addressing unconscious bias technology, data and science can help further.
So is AI the answer?
With Covid-19 seemingly exposing long-standing societal divides, discussions surrounding AI and the role it could potentially play in improving Equality, Diversity and Inclusion have come back to the forefront. Over the last 12 months there has been a dramatic shift in vacancy positioning and the power of the employer brand, with Work From Home rising as a search term, flexible hours becoming the norm and an increase in more social interaction to display company culture. As Equality, Diversity and Inclusion is trending highly for being on the agenda of all HR and Talent Acquisition professionals for 2021, there has also been an increase in interest in AI-powered recruiting applications to deliver a cost and time efficient solution, whilst mitigating human bias for a fairer overall hiring process.
Like any new technology, artificial intelligence is capable of immensely good outcomes, but in balance there are some negatives. There is a lot of focus and fear surrounding AI and shift in human control to empowering technology. However, AI is simply the simulation of human processes by a machine. Argument suggests that the deepest-rooted source of bias in AI is the human behaviour it is simulating; a biased data set used to train the algorithm. Yet we have come along way since the infamous Amazon AI recruitment blunder and there have been countless other businesses initiating AI processes to support their commitment to developing an inclusive workforce. Artificial intelligence is taking its learnings from humans, but it is still technology and technology is neutral. If AI-assisted recruitment initiatives are carefully crafted and validated, it does offer a solution for mitigating human bias in the hiring process.
Understanding where bias could be addressed in your hiring process
Typically, hiring is not a single decision, but a collective set of decisions involving cross-functional teams and different recruitment stages. Unconscious bias may be influential at various stages in the hiring process where human judgement or decision making occurs, such as in the advertising when sourcing candidates, the criteria used for sifting and scoring.
- Optimised sourcing
Your sourcing strategy combines a mix of designing attractive job descriptions, posting various advertisements, headhunting and social campaigns. The language, visual imagery and location used across these promoted vacancies will impact the diversity of your candidates. Beyond making conscious decisions to mitigate this or working with your recruitment software partner, using artificial intelligence can further optimise the display of job ads, including neutralising the job description content to ensure you are appealing to a wider, more diverse audience.
- Intelligent sifting
A crucial part of the recruitment process, and especially challenging when there is a large volume of applicants, is your screening and sifting process. Utilising a savvy Applicant Tracking System can provide great efficiency through automated processes, but implementing AI-assisted recruitment screening is the next step on, using algorithms systematically decipher the submitted CV or application and appraise vital information against scoring criteria, such as years of experience, languages, and qualifications. AI-assisted recruitment is one step beyond blind CVs and applications, which offer an immediate support for your diversity strategy. AI, however, can be an efficient and cost effective way to narrow your talent selection, whilst remaining unbiased as the technology grades purely on factual data with no human influence.
- Recruiter chatbots
As part of the interview process, some organisations are trialling the use of chatbots, to provide real-time interaction with candidates by asking relevant role-based questions and providing company insight, next steps and feedback. This strategy is a hybrid stage between sifting and interview and not only streamlines the decision-making process, but improves the candidate experience too.There are drawbacks to utilising chatbots, such as language barriers and greater challenges with spamming and inappropriate candidate interaction, but this could be one to watch for the future.
- Virtual interviews
In the last 12 months, all organisations have been forced to embrace technology and take interviews online, however AI can take these interviews one step further. AI technology has the power to use facial expression and language use analysis to monitor responses and reactions during video interviews. This can be used to identify key skills and personality traits, to further aid the decision making process and adds a further level of discrimination free insight beyond using recorded or live video interviews.
AI-assisted recruitment can offer greater efficiencies and enable you and your team to remain unbiased throughout some stages of the recruitment process, but it is not necessarily an independant or easy solution.
AI is not a tick box exercise
One of the biggest drawbacks of integrating AI-assisted recruitment is that the tools are often viewed as a quick fix solution. However, as Equality, Diversity and Inclusion is a complex challenge and every organisation has their own nuances to their process, policies and approach to ED&I, so the solution has to be detailed and meticulous. This takes great company-wide buy-in. Growing a diverse workforce goes well beyond recruitment and it is not something organisations can afford to just pay lip service too. Recruitment is often just the first step in the wider strategy.
Organisations that are successfully implementing AI-assisted recruitment have found that continuing to run traditional practices, almost running their own study, has supported wider buy-in and opened doors to the positive possibilities of utilising AI to drive a more inclusive working environment. Yet it is a continual learning process and as technology evolves, the tools we utilise to support recruitment, such as AI, will need to evolve too.
Weighing up the benefits
The benefits of implementing AI-assisted recruitment are vast, from budget management, through to researched evidence of it reducing bias in the recruitment process to deliver a more diverse talent pool for opportunities.
- Maximising efficiencies
One of the key original drivers for utilising AI-assisted recruitment is the time and cost savings associated with its implementation. AI can quickly assess an entire pipeline of candidates and follow automated processing, so by the time the candidates reach any human interaction, the volume is reduced and done so without bias. This has been especially pertinent in the last 12 months when the volume of applicants has dramatically increased due to the Covid climate. Although AI offers partial judgement, often even implementing an automated process alongside your ATS can offer a great step forward in improving efficiencies.
- Designed to suit your needs
The beauty of AI-assisted recruitment is that it can be designed to meet the specific criteria you need to execute your organisation’s strategy. There are AI practitioners and movements such as OpenAI putting together principles for ethical and fair AI, but the possibilities are vast beyond this best practice framework.
- Continual evolution
As with all technology, AI will continue to evolve to offer even greater efficiency and unbiased processing to support your Equality, Diversity and Inclusion strategy. AI can be built to be tested for bias, with detailed validation and rigorous testing of the AI algorithm, so it will be possible long-term to integrate and evolve your processes in tandem with your AI-assisted recruitment tools.
Integrating AI into your Equality, Diversity and Inclusion strategy
Embracing AI-assisted recruitment tools provides us with the opportunity to make decisions grounded in unbiased data. You and your organisation will still be the one making the decision, but these will be informed by data driven free from bias. It is impossible to eradicate unconscious human bias and AI-assisted recruitment offers an exciting, innovative solution. Despite the fear and media stories surrounding poorly tested or implemented AI strategies for driving diversity in recruitment, let us not forget the bias that currently exists in traditional recruitment practices. Although steps can be taken and this last year has seen a rise in the role of automation, blind shortlisting and greater awareness around language within job descriptions, AI assisted recruitment could take it one step further.
The benefits are clear and if we take critical steps to adopt best practices and implement validation testing, we can look to truly harness this technology to diversify the workplace.
For more information on how recruitment technology can innovate your recruitment and help you to pioneer your ED&I strategy, contact networx today.Back to blog page