Artificial intelligence (AI) is increasingly pervasive in HR industry and brings lots of potentials. This article is focused on deblackboxing the ways that Artificial Intelligence change the HR industry and trying to find out its advantages and limitations through analyzing the design principle and algorithm, and then provide business advices of future AI use in HR industry.
Nowadays, AI and ML technologies are advancing at a phenomenal rate and immersing into different kinds of industries. To some extent, this is a world of “ubiquitous computing’. The English Oxford dictionary defines AI as “The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.” Three core components — high-speed computation, a huge amount of quality data and advanced algorithms differentiate AI from ordinary software (The new age: artificial intelligence for human recourse opportunities and functions, 2018).
AI technologies offer great opportunities to business. More than 40% of employers across the globe already use AI in some way, a Deloitte study in 2018 found, with “exponential” growth expected over the next five years. One of the changes brought by AI is improving HR functions, such as recruiting and talent acquisition, providing real-time information to employees, training services, etc. According to Forbes Human Recourse, AI may revolutionize and redefine the recruiting and hiring process. (The Rise of AI In HR: Nine Notable Developments That Will Impact Recruiting And Hiring, 2018) Not only the employment-oriented service business, like LinkedIn, many other companies — from Allstate to Hilton to Five Guys Burgers and Fries — are also using AI to help recruiters and hiring manager screen, review and interview job candidates.
This article is focused on deblackboxing the ways that Artificial Intelligence change the HR industry and trying to find out its advantages and limitations through analyzing the design principle and algorithm. The purpose of the article is to explain AI application in HR industry to the business and ordinary people, and then provide business recommendations when using AI system. The entire article could be divided into three main parts: 1) introduce the ways that AI applied in HR industries and its technical components; 2) present the limitations and challenges of AI development in HR industry; 3) provide business advices of future AI use in HR industry.
The ways that AI applied in HR industries
In general, there are main three ways that AI can be applied in HR industries, including staff forecasting, hiring talents and personalizing employee experience. Then, I will introduce the three ways separately in the following article.
2.1 Staff Forecasting
Machine learning and prediction is possible because the world has regularities. Things in the world change smoothly. We are not beamed from point A to point B, but we need to pass through a sequence of intermediate locations (Machine Learning Chapter 2&3, 2017). Today, utilizing machine learning to predict human behavior has become very commonplace. Through learning from huge amounts of data, we are able to get a general model or pattern, and then complete some prediction.
Take employee’s turnover prediction as an example. There is no doubt that every year, the company need to face inevitable turnover and attrition. Now AI is able to inform the HR department of their employee’s decision before they leave the company, so that the company could make some preparation in advance. Using employee engagement data, whether it be from pulse surveys, brand advocacy or performance gamification, artificial intelligence could determine an employee’s level of interest, match the current model that made by previous training data, and then give a prediction on whether they are trying to change positions.
In addition, a stronger digital IQ will bring a business deeper into what is referred to as an “unconscious level” of information. ( (The new age: artificial intelligence for human resource opportunities and functions, 2018) By analyzing people’s statements, mood and intentions on social media, along with other public- data sources, human behavior can be simulated by autonomously learning machines. This makes it possible to validate the employee experience on a day-to-day basis. HR performance and succession data provide information on which employees are engaged and challenged. That gives a new dimension to strategic workforce planning to reduce employee attrition.
Not only predict whether the staff decide to leave company, AI can also predict the following important things (How Artificial Intelligence Impacts HR: Incorporating intelligent assistants in the workplace, 2016)
- Which job candidates will make the best hires
- Which employees are most likely to leave the organization
- What kind of compensation packages are most likely to lead to employee retention
- What the need and availability will be for employees with certain skill sets
- Which benefit packages are most likely to appeal to employees
However, AI prediction cannot be used in the the HR departments of small companies, because in problems where data are limited, machine learning often is not an ideal solution (Deep Learning: A Critical Appraisal). Human beings can learn abstract relationships in a few trials, but machine learning needs thousands, millions or even billions of explicit training examples. The employee’s digital data in those small companies is not enough to implement deep learning.
2.2 Hiring Talents
Selecting talents from huge amounts of resumes is a really tiring and tedious work. Owing to lack of time and energy, the interviewers usually take less than a minute to decide whether an interviewee is an ideal match for the job or not. Based on the candidate’s appearance, speech, experience and the way they present themselves, random decision to hire people is made without the help of data, so HR department are very likely to make a wrong selection. According to ‘Artificial Intelligence in HR’, hiring managers complain about getting 30 to 40 percent of their candidates’ wrong.
Nowadays, Artificial Intelligence is able to free HR staff from boring resume scanning work by utilizing Natural Language Processing. Natural Language Processing (NLP) is a wide area of research where the worlds of artificial intelligence, computer science, and linguistics collide. It includes a bevy of interesting topics with cool real-world applications, like named entity recognition, machine translation or machine question answering. Each of these topics has its own way of dealing with textual data.
Take us take Google Hire to see how AI to scan resumes by Natural Language Processing. Launched in 2018, Google Hire is an applicant tracking system developed by Google that helps small to medium businesses (SMBs) to distribute jobs, identify and attract candidates, build strong relationships with candidates, and efficiently manage the interview process (Wikipedia). Machine learning makes it possible to identify the skill and capabilities required for a certain open position. Based on the job requirements, the app will select the strongest matches after analyzing each candidate’s skills, education, location, salary preferences, etc. This capability doesn’t have to be restricted to the pool of individuals who have applied for a specific job. The search could be extended to individuals who have previously submitted their resume or who are regularly looking for new positions within the organization.
Besides, an AI system could even help company predict a certain level of culture fit based on those attributes or other skills that the prospect listed on their resume. Among other processes, some AI platforms could also schedule interviews by finding optimal availability for all parties involved.
According to Business Insider, about 90% of big companies in America use AI system to scans resumes for keywords and then forwards on “qualified” candidates to the company’s HR team. In a market where you’ve got hundreds, and even thousands, of applicants for a single job, it’s the most efficient use of a company’s time.
2.3 Personalizing Employee Experience
Artificial intelligence has undergone a paradigm shift from logic-based to interactive (agent-oriented) models paralleling that in software engineering. Interactive models provide a common framework for agent-oriented AI, software engineering, and system architecture (Why Interaction Is More Powerful Than Algorithms).
Artificial Intelligence in HR industry could support employees to find the right information, with lower costs, in less time and in a secure manner. Every day, the HR staff need to deal with different kinds of questions from the employees, so sometimes they are unable to provide very patient and nice solution for everyone. Besides, some new employees are not familiar the departments in the company and puzzled with whom to ask for help. According to a report by Deloitte, nearly 80 percent of executives rated employee experience very important or important, but only 22 percent reported that their companies were excellent at building a differentiated employee experience (EX) (How AI Chatbots Can Help Transform The Employee Experience, 2014). The conversational AI solution uses a machine learning capability — Natural Language Processing can solve all the problems mentioned above.
Humans and learning machines are working together to produce an ever-increasing amount of HR data in the computing cloud, and the use of artificial intelligence analyses can automatically offer answer based on the cloud database to the employee question. AI will help to efficiently automate many back-office functions for reliable HR transactions and service delivery. This document is focused on conversational AI capabilities for HR transactions and provides insight about intelligent automation via the technology-agnostic Chatbot (The new age: artificial intelligence for human resource opportunities and functions, 2018).
Source: The new age: artificial intelligence for human resource opportunities and functions
The above diagram depicts a high-level technology landscape for an HR solution. It shows how to access and maintain HR transactions, via Chatbots, using conversational AI.
For new employees, conversational AI for the HR system will help get them up to speed fast. For example, an AI-powered program could provide the names, locations and contact information for people they should connect with during their first week. New employees could also be advised by AI engines to check out a new-hire web page containing useful information, including training modules and business-conduct guidelines. For department managers, they can also utilize conversational AI to schedule a meeting and book a meeting room, or even check their employees’ performance within one week. Additionally, chatbots also send personalized notifications to employees about company policies, rewards, holidays and so on. This way, your employees are kept informed about the latest updates without having to access or navigate the mails to find them. The real-time chatbot can significantly improve people’s working efficiency.
Limitations and Challenges
Artificial intelligence is increasingly influence HR industry, but there is a growing awareness of the effect of bias problem and data privacy in machine learning. In this section, we will discuss them separately.
3.1 Bias in Artificial Intelligence
Last year, Amazon.com Inc’s (AMZN.O) machine-learning specialists uncovered a big problem: its new system was not rating candidates for software developer jobs and other technical posts in a gender-neutral way. That is because Amazon’s computer models were trained to vet applicants by observing patterns in resumes submitted to the company over a 10-year period. Most came from men, a reflection of male dominance across the tech industry, just like the following image:
Source: Latest data available from the companies, since 2017. By Han Huang | REUTERS GRAPHICS
Amazon tries to revise the the program, but that was no guarantee that the machines would not devise other ways of sorting candidates that could prove discriminatory.
Every AI system is dedicated to map human mind and the reality into binary structure and computation system. Actually, this is not the only case that found AI has bias on gender, races, class, etc. Over centuries, humans developed critical theory to inform decisions and avoid basing them solely on personal experience. However, machine intelligence learns primarily from observing data that it is presented with. While a machine’s ability to process large volumes of data may address this in part, if that data is laden with stereotypical concepts of gender, the resulting application of the technology will perpetuate this bias. (Gender Bias in Artificial Intelligence: The Need for Diversity and Gender Theory in Machine Learning, 2018)
3.2 Data Privacy
Deep learning, as it is primarily used, is essentially a statistical technique for classifying patterns, based on sample data, using neural networks with multiple layers. In problems where data are limited, deep learning often is not an ideal solution. (Deep Learning: A Critical Appraisal) Therefore, AI application in HR industry requires huge amounts of human data to operate, including the data from questionnaire, business file, interaction with Chatbots, resume, or even their personal social media. To some extent, employee are surveilled by their company all the time.
Now some companies have made relative rules to protect people’s data privacy, such as not collecting Personally identified information (PII), which was defined in OMB Memorandum M-07-1616 refers to information that can be used to distinguish or trace an individual’s identity. It includes our name, personal identification number, address information, personal facial characteristics, etc. The table below is the DPI (Department of Public Instruction) PII examples (not all inclusive).
Nowadays, technology is developing so fast that law regulation could not catch up its development speed and always lag out. Our PII is very valuable asset and it belongs to us, but some companies are collecting and tracking our personal data and selling our data without our consent and knowledge. The rules of Internet privacy could not just be conducted by those Internet giant. This principle prescribes that any matter which is essential because it either concerns fundamental rights of individuals or is important to the state must be dealt with by a parliamentary, democratically legitimized law. (Paul Nemitz)
Through the article, we could conclude that artificial intelligence really bring some changes in HR industry, including staff forecasting, hiring talents and personalizing employee experience. By learning from huge amounts of data in the company cloud dataset, HR managers are able to get a general model or pattern, and then do some prediction, such as the turnover of the employees, the packages which are most likely to appeal to employees, etc. Artificial Intelligence is also able to free HR staff from boring resume scanning work by utilizing Natural Language Processing. Using Artificial Intelligence for resume scanning can not only target the employees who are most fit the companies’ requirements, but also save time of HR staff so that they can make more effort on human service work. Besides, Artificial Intelligence in HR industry could support employees to find the right information, with lower costs, in less time and in a secure manner by using Chatbots. The conversational AI solution uses a machine learning capability — Natural Language Processing can personalize workers’ user experience and make working easier and more efficient.
However, the world is moving faster with new technologies, and it is easy for organizations to make missteps. Artificial intelligence is not the perfect solution, and it also has lots of challenges and limitations. The major problems are data privacy and bias issue. If that data is laden with stereotypical concepts of gender, the resulting application of the technology will perpetuate this bias. In addition, some companies use employee’s data without any consent. Law regulation could not catch up its development speed and always lag out. We could also find that it is sometimes hard to use AI in small companies, because there are not enough training data to be fed into the AI system.
My advice to the company using AI in HR industry is listed below:
1、Eliminate bias. AI applications are capable of processing speed, but they can go wrong due to biased learning input. An AI solution can be a catalyst for positive change if it has been used in the correct way.
2、Data protection. Companies have the responsibility to protect their employee data privacy. All of the regulation and procedure of using employees’ data must be very transparent and democratic.
3、Keep moderate expectations, with a scalable solution. Don’t make business decision only based on artificial intelligence. Human participation must be involved in the AI system.
AI-based HR applications have strong potential to raise employee productivity, but we also need to be cautious to use it.
- Aimee, G. (2012). Most big companies have A tracking system that scans your resume for keywords., 2019, from https://www.businessinsider.com/most-big-companies-have-a-tracking-system-that-scans-your-resume-for-keywords-2012-1
- Alpaydin, E. (2016). Machine learning : The new AI. Cambridge, Massachusetts: The MIT Press.
- Aurora. (2017). Featured research summary – the state of artificial intelligence in HR. HR Strategy and Planning Excellence Essentials,
- Bhatia, T. (2018). Artificial intelligence in HR. HR Strategy and Planning Excellence Essentials,
- Jagan, J. (2018). How AI chatbots can help transform the employee experience?, 2019, from https://aibusiness.com/ai-chatbots-employee-experience/
- Light, R. (2017). How artificial intelligence will revolutionize HR: HR’s job is safe. Core HR, HRIS and Payroll Excellence Essentials,
- May, R. (2016). How artificial intelligence impacts HR: Incorporating intelligent assistants in the workplace. HR Strategy and Planning Excellence Essentials,
- Sheila, M. (2018). The new age: Artificial intelligence for human resource opportunities and functions., 2019, from https://www.ey.com/Publication/vwLUAssets/EY-the-new-age-artificial-intelligence-for-human-resource-opportunities-and-functions/$FILE/EY-the-new-age-artificial-intelligence-for-human-resource-opportunities-and-functions.pdf
- Yarosalv, K. (2018). How AI is revolutionizing applicant tracking and the recruitment process?, 2019, from https://www.hrtechnologist.com/articles/recruitment-onboarding/how-ai-is-revolutionizing-applicant-tracking-and-the-recruitment-process/