‘Alexa, give my Assistant a 5% raise’
‘Hey Siri, tell the Office Manager how many days holiday they have left’
‘Ok Google, where can I download our Employee Handbook from?’
With Artificial Intelligence (AI) becoming more popular and widely used in our personal lives, Joe Taylor and Daniel Gray examine how AI can benefit Human Resources (HR) functions, and what employers should consider when implementing AI in HR. But first, what is AI?
What is AI?
AI is software that undertakes pre-assigned tasks and makes decisions and predictions depending on what it has been ‘taught’. The more a particular piece of AI is used, the more it ‘learns’ but in the first instance, an AI user ‘tells’ the AI what environment it operates in and what the desired outcomes are. This usually takes place by machine learning based on collected data and data input. For example, if you want AI to automatically delete all emails received from colleague A, first you have to tell the AI colleague A’s name and email address, so the AI knows what to ‘look’ for when ‘deciding’ what to delete. The AI will not automatically be able to do this but, once you have ‘told’ it this information, it will then action this.
How can AI be used by a HR function?
In the HR sphere, there are various uses for AI. We focus in this article on three areas where it can be advantageous to have such systems in place – Recruitment, Employee Engagement, and Learning and Development.
Recruitment
There are a variety of different AI systems on the market that can assist with the recruitment process. For example, AI can ‘read’ CVs and make decisions on whether to progress an application based on the information in the CV, applying the parameters the AI has been ‘taught’ to look for. Ideally such decisions can be overridden, and should be monitored by HR staff, but an initial first sift of job applications and CVs by AI could save HR teams hundreds of hours a year.
Further, chatbots are increasingly used at the first stage of the interview process, with applicants answering questions posed by an AI in a video call, with, in some instances, the AI deciding whether to progress their application after such interview.
AI can also benefit potential job applicants during the recruitment process. At each stage, there may be some duplication of information that must be provided but an AI can streamline the application process and design more user-friendly and intuitive (and less repetitive) forms.
For an article that looks at AI in recruitment and the risk of discrimination in using such tools, click here.
Employee Engagement
HR teams’ time can be further saved with the assistance of AI with employee engagement. For example, having an internal chatbot function which allows employees to ask it frequently asked questions (the answers to which have been prepopulated) will save busy HR teams having to repeat themselves and duplicate work – if an employee asks a chatbot to tell them where they can find a copy of their employer’s privacy notice, the chatbot ought to be able to direct the employee to it, rather than a member of HR staff having to do the same.
AI can assist with administrative or repetitive tasks, as referred to above. This will, ultimately, reduce overheads and staffing costs as well as giving HR professionals more time to focus on other areas.
Learning and Development
Finally, AI can also be used to tailor and recommend an employee’s development areas. By monitoring employee searches and output, as well as understanding their job description and skillset, a suitable AI will be able to assess areas for improvement and identify an employee’s knowledge gaps. From an employer's perspective this can help demonstrate that you are taking an active interest in your employee’s progression and development, fostering strong employer-employee relations which will benefit your business.
For information on how algorithms can be used in the workplace, click here.
Employer considerations
Whilst there are some advantages from using AI in the HR function, before deciding what and how to use it, employers should consider the following:
- Unconscious bias and discrimination: AI will, initially at least, ‘learn’ from input data (which may be based on current or historical human bias). Therefore, HR teams should bear in mind that if the data they are using to ‘train’ the AI contains bias, this will likely be reflected in the decisions and results the AI makes. If HR teams rely on an AI decision which is influenced by bias, this could result in allegations of discrimination, which could lead to actionable claims by job applicants. HR teams should therefore be aware of this risk and thoroughly test any AI both prior to using it with candidates, and routinely afterwards, to monitor any impact the AI may have on the diversity of successful candidates and mitigate the discrimination risks.
- Sensitive personal data: HR data typically includes, for the purposes of UK GDPR, ‘special category data’, which by its nature is sensitive personal data to which additional data protection safeguards apply. This places greater obligations on employers from a data processing perspective. When involving AI with such special category data, consideration should be given to how that data is processed, stored and secured, to aid compliance with an employer’s data protection obligations. The EU's incoming AI Act has classed HR AI as ‘high risk.’ To find out more about what this means and the draft AI regulations themselves, click here.
- Employee relations: If AI is overused, for example in the recruitment process or with employee engagement, the employee user experience may begin to feel impersonal. Identifying a blended approach where AI can best be used within an HR context will be key. Further, AI at present is unlikely to be able to read emotions, understand the nuances of how different teams function on a personal level, or understand how different employees react and respond to information. Thought should be given to managing the use of AI in situations where the AI’s response, whilst factually accurate, may not be the most appropriate.
- AI is still early stage: Whilst there are some very sophisticated AIs tools on the market, AI is still in its relative infancy. When ‘teaching’ AI there will likely be some teething problems, which HR teams and employers more generally should be live to when implementing AI solutions within the business.
Conclusion
Notwithstanding the above considerations and potential risks to employers, using AI in the HR sphere has some clear benefits. In addition to those outlined above, AI can identify issues and trends within a business, helping employers to avoid and/manage them in a timely fashion. Employers may therefore wish to consider whether and how they might introduce some AI to assist the HR function, as well as exactly where within the HR function such AI could be deployed to best effect. If you would like to discuss this with us further, do get in touch.