The Good, the bad and the ugly of AI and machine learning in recruitment

Recruitment has undergone a significant transformation over the years. The traditional hiring methods have been replaced with the use of technology, which has made the process more efficient and effective. One of the technologies that have gained popularity in recruitment is Artificial Intelligence (AI) and Machine Learning (ML). Artificial intelligence (AI) and machine learning (ML) have become important tools for recruiters looking to streamline their hiring process. AI and ML can be used to quickly assess large numbers of applications, identify key skills and qualifications required for roles, and even predict the success rate of candidates in a certain role. In addition, AI and ML can help recruiters discover potential candidates who may otherwise go unnoticed.

In recent years, the use of AI and ML in recruiting has grown significantly. This comes as no surprise, given that the worldwide Artificial Intelligence (AI) market was valued at US$95.9 billion in 2022 and is expected to expand at a CAGR of 32.8% during the forecast period, resulting in a market size of US$276.3 billion by 2026, according to research conducted by Global Industry Analysts Inc. In our podcast, The lasting legacy of talent trends: Separating fads from reality, HR expert Soreya Senior and our CEO, Sam Ingram, share valuable insights on this topic.

Considering the increasing use of AI and ML in recruitment processes, it’s clear that this technology offers many potential benefits for employers – including quicker decision-making, improved candidate matching, increased automation, and enhanced accuracy. As more organisations adopt AI and ML technologies for their recruitment processes, it’s essential that they understand how best to use them.

For example, employers should ensure that their algorithms are unbiased so as not to disadvantage certain groups. They should also consider what data sources and metrics they should use when training their models – such as past performance ratings or education level – as well as how often they need to retrain these models to remain effective. It’s important that organisations approach this technology responsibly so as not to put vulnerable groups at a disadvantage or create an environment less conducive to diversity hiring practices.

In this blog, we will explore the good, the bad, and the ugly sides of AI and ML in recruitment and address some key strategies that can help mitigate the risk associated with AI and ML-based recruitment processes.

Advantages of AI/ML-based recruitment processes

• Faster decision-making and improved candidate-matching capabilities: With AI and ML technologies, recruiters can make faster decisions based on the data gathered, which helps them to quickly identify and select the best talent for job openings. The technology also enables increased automation of recruitment processes such as candidate screening and assessment, thus improving the accuracy and efficiency of the process.

• Automation of repetitive tasks: AI and ML can be used to automate repetitive and time-consuming tasks in recruitment, such as CV screening, scheduling interviews, and sending follow-up emails. According to a survey conducted by LinkedIn, 67% of talent acquisition professionals believe that AI and ML can save them time by automating administrative tasks.

• Increased automation of recruitment processes such as screening and assessment: Machine learning models can be used to enhance the accuracy of predictions regarding which candidates are most likely to succeed in a given role. By providing insights into a candidate’s skillset and work history, these models help recruiters make accurate hiring decisions that are tailored to their specific needs.

• Enhanced accuracy of predictions regarding successful candidates: AI-driven software can be utilised to find qualified candidates more quickly than ever before by scouring social media platforms – like LinkedIn – for potential matches that fit a job description. This type of software can rapidly scan through profiles and sift through vast amounts of user data online in an effort to identify promising candidates with just the right skills for an opening.

• Better candidate experience: AI and ML can also improve the candidate experience by providing quick and personalised responses to candidates. Chatbots, powered by AI, can answer candidates’ questions, schedule interviews, and provide feedback on their application status. This improves candidate engagement and reduces the time-to-hire.

• Cost reduction: AI and ML can reduce the cost of recruitment by automating tasks and reducing the need for human intervention. According to a study conducted by Korn Ferry, a global organisational consulting firm, AI can reduce the cost-per-hire by up to 70%.

Disadvantages of AI/ML-based recruitment processes

• Risk of bias: One potential disadvantage of using AI and ML technologies in recruiting is the risk of bias creeping into decision-making because of automated processes. For example, if the algorithms used to identify potential candidates for a job opening are trained on a biased data set, then it can lead to some qualified applicants being overlooked or even blocked from consideration altogether. According to a study by the Harvard Business Review, Amazon had to scrap its AI recruiting tool because it was biased against women.

• Algorithmic errors: Additionally, the use of AI in recruitment can lead to algorithmic errors that result in incorrect decisions being made which can unfairly harm individual applicants. This could range from mistakes like sending out generic rejection emails to applicants who did meet the requirements for an opening or falsely labelling certain candidates as qualified when they are not truly suited for the role. Also, AI and ML algorithms require large amounts of data to be effective. If the data used is incomplete, inaccurate, or outdated, the algorithms will produce inaccurate results. This can lead to wrong hiring decisions and damage the reputation of the company.

• Inability to assess qualitative traits: Furthermore, there is also a risk that AI-based recruitment systems might not be able to accurately assess certain qualitative traits like soft skills or communication abilities when making selections, thus leaving out potentially great candidates who may have otherwise been hired had they had been interviewed by recruiters in person.

• Privacy concerns: Privacy concerns can also arise with the use of AI and ML technologies as these systems collect huge amounts of personal data from applicants which could be misused by malicious third parties if not properly secured. Therefore, it is important for companies implementing automated recruitment processes to take measures such as encrypting all collected data and investing in robust cybersecurity protocols in order to protect users’ information from theft or abuse. Forbes warns that companies should be aware of potential privacy issues when using AI in recruitment processes; if personal data is mishandled during these processes then it could lead to costly fines or reputational damage. Companies should develop secure policies around collecting personal information from applicants prior to implementing any automated systems as part of the recruitment process. Furthermore, organisations should make sure they only collect relevant personal data which is necessary for assessing candidates’ qualifications; anything else should be considered superfluous and avoided accordingly.

• Lack of human touch: The relevance of human touch and relationship building has always remained pivotal for our CEO Sam Ingram who has often stressed on the need for human relationship and highlighted how that is what acts as the differentiating factor from your competition. Listen to Sam as he talks about the importance of human interaction in a technology-dominated world.

“I think the one thing we spoke about earlier this year is that you still must have human relationships. I think all this technology is wonderful in a world that is growing and growing quickly, and the demand is massive. But what tends to get you through from a business perspective is human communication and interaction. Keep it simple, stupid. Just communicate better.” – Sam Ingram, CEO Northreach

While AI and ML can automate certain tasks, they cannot replace the human touch in recruitment. Candidates still want to interact with human recruiters, and AI and ML cannot replace the emotional intelligence and empathy of human recruiters.

• Lack of diversity: When AI and ML algorithms are trained on historical data that reflects bias and discrimination, it can perpetuate this bias into the future, leading to a lack of diversity in the workplace. Moreover, AI and ML algorithms are often “black boxes,” meaning that it is difficult to understand how they make decisions. This lack of transparency can make it difficult to detect and correct bias in the algorithm.

How to best use AI/ML-based recruitment processes

Here are some strategies for utilising AI and ML-based recruitment processes in the most effective and efficient way possible:

1. Establish a clear set of objectives: Defining goals and expectations upfront can help ensure that the chosen recruitment methods will meet the organisation’s needs.
2. Incorporate human evaluation into decision-making: While algorithms can be helpful for sorting through large amounts of data, humans should still play an important role in evaluating applicants and making decisions about which candidates to move forward with at each stage of the process.
3. Assess potential risks: Using AI and ML-based approaches may increase the risk of introducing bias into decision-making, algorithmic errors leading to incorrect decisions, or failing to accurately assess certain qualitative traits in applicants. Taking steps to mitigate these risks may include setting up regular auditing processes and reviewing algorithm performance regularly.
4. Focus on user privacy: Consider what types of personal data is being collected from users as part of the recruitment process, how it is being stored and used, who has access to it, and what measures have been put in place to protect user privacy.
5. Utilise feedback loops: Continuous improvement is key when using any type of automated system. Leveraging feedback loops to adjust parameters and improve performance over time can help ensure that your AI/ML systems are getting better at providing valuable insights into job applicants over time as well as staying compliant with laws and regulations governing their use.

AI and ML technologies in recruitment

There are several AI/ML-based recruitment applications available in the market today that offer a range of features to automate and streamline the recruitment process. Here are some examples:

1. HireVue: This platform uses AI-powered video interviews to assess candidates’ skills and personality traits. It analysis facial expressions, tone of voice, and word choice to provide insights into the candidate’s suitability for the role.
2. Mya: This conversational AI chatbot engages with candidates to answer their queries, schedule interviews, and provide feedback. It also screens CVs and job applications using ML algorithms to identify the best candidates.
3. Textio: This platform uses natural language processing (NLP) and ML to analyse job descriptions and provide suggestions for language that will attract a diverse pool of candidates. It can also analyse candidate communications to identify potential biases.
4. Eightfold: This platform uses ML to match candidates to job opportunities based on their skills, experience, and interests. It also offers personalised career development recommendations and training resources.
5. Entelo: This platform uses ML algorithms to analyse candidate data from multiple sources, such as social media, job boards, and employee referrals. It provides insights into the candidate’s experience, skills, and suitability for the role.

Conclusion

Overall, AI and ML technologies have revolutionised the way recruiters approach their work by providing them with powerful tools they can use to streamline their hiring processes and find ideal candidates more quickly than ever before. In conclusion, AI and ML-based recruitment processes can be effective tools for streamlining the hiring process and providing insights into job applicants that may not be available through other means.

However, organisations should consider potential risks associated with using automated systems and should develop strategies for mitigating those risks while still taking advantage of the capabilities offered by these technologies. Additionally, implementing feedback loops to continuously improve system performance is essential in achieving long-term success when using AI/ML-based recruitment processes.

Key Takeaways

• Establish a clear set of objectives before implementing AI/ML-based recruitment processes.
• Incorporate human evaluation into decision-making.
• Assess potential risks associated with using automated systems.
• Focus on user privacy when collecting personal data.
• Utilise feedback loops to adjust parameters and improve performance over time.