Artificial Intelligence (AI) in Re- cruitment
Essi Rankila
BACHELOR’S THESIS November 2020
Degree Programme in International Business
ABSTRACT
Tampereen ammattikorkeakoulu
Tampere University of Applied Sciences Degree Programme in International Business ESSI RANKILA
Artificial Intelligence (AI) in Recruitment
Bachelor's thesis 59 pages, appendices 8 pages November 2020
Artificial intelligence causes discussion in many industries and recruitment is one of the fields where development steps are taken. The utilization of artificial intel- ligence is truly a relevant matter in the recruitment industry. The rapid develop- ment and countless opportunities provided by artificial intelligence has increased the interest towards it. This thesis was not done as a cooperation with a commis- sioner.
The aim of this research was to map out the current use of artificial intelligence in Finnish personnel companies’ recruitment, and to predict how the use of artifi- cial intelligence in recruitment will evolve. The research will also give an insight on what value artificial intelligence brings to recruitment.
The empirical part of the research was conducted as a quantitative research, but the research has also qualitative features. The data were collected by sending an online survey to Finnish personnel service companies that are members of the Private Employment Agencies Association in Finland. The survey was conducted in October 2020.
The results revealed that the utilization of artificial intelligence in Finnish person- nel service companies’ recruitment process is still quite low. The current use of AI in recruitment is mostly in applicant reviewing and preselection, applicant com- munication, and customer service. The opportunities provided by AI will be more utilized in the future once AI is more evolved. The value of artificial intelligence in recruitment is seen in the faster recruitment process, and replacement of the time-consuming routine tasks, enabling more time for the company’s core busi- ness. In the future artificial intelligence is hoped to better combine the work and the employee.
Further research may be needed to truly understand the current state of artificial intelligence in Finnish personnel service companies’ recruitment. A supplemen- tary research about the same topic in five years would be beneficial as the current research also studied the future of AI in recruitment.
artificial intelligence, AI, recruitment process outsourcing, recruitment process
CONTENTS
1 INTRODUCTION ... 6
1.1 Research background ... 6
1.2 Thesis objective, research questions and limitations ... 6
1.3 Theoretical framework ... 8
1.4 Research methodology and data collection ... 8
1.5 Thesis structure ... 9
2 RECRUITMENT PROCESS OUTSOURCING ... 11
2.1 Recruitment process outsourcing ... 12
2.1.1 Assignment and initiating the recruitment ... 12
2.1.2 Interview process and selection ... 14
2.1.3 Interview questions and methods ... 14
2.1.4 Selection, billing and aftercare ... 15
3 ARTIFICIAL INTELLIGENCE IN RECRUITMENT ... 17
3.1 Artificial intelligence ... 17
3.2 Utilization of AI in recruitment ... 18
3.2.1 Job application processing and matching ... 19
3.2.2 Applicant communication and customer service ... 19
3.2.3 Job advertisement keyword planning and targeting... 20
3.2.4 Interviewing ... 21
3.3 Challenges of AI in recruitment ... 21
3.4 Future of AI in recruitment ... 22
4 RESULTS AND DATA ANALYSIS ... 25
4.1 Research methods and data collection ... 25
4.2 Results ... 26
4.2.1 Respondent background ... 27
4.2.2 What kind of image do you have about artificial intelligence in general? ... 29
4.2.3 Is artificial intelligence currently being utilized in your company’s recruitment process? ... 30
4.2.4 Does your company have any plans to introduce artificial intelligence in the recruitment process in the future? ... 31
4.2.5 In which areas of recruitment is artificial intelligence currently utilized in your company?... 32
4.2.6 Do you feel that artificial intelligence has a positive impact on recruitment? ... 34
4.2.7 Do you feel that artificial intelligence has changed recruitment? ... 36
4.2.8 Do you feel that artificial intelligence is going to change
recruiting in the future? ... 38
4.2.9 In which areas of recruitment do you find artificial intelligence to be the most useful? ... 40
4.2.10Briefly describe how your company, your client companies and your job seekers will benefit from the effects of artificial intelligence ... 41
4.2.11Do you have anything else to add? ... 42
4.3 Analysis of the research material ... 42
5 DISCUSSION AND CONCLUSION ... 45
5.1 Validity and reliability ... 46
5.2 Evaluation of the thesis process and suggestions for further research ... 47
REFERENCES ... 49
APPENDICES ... 52
Appendix 1. Online survey ... 52
Appendix 2. Cover letter ... 58
Appendix 3. Reminder email ... 59
ABBREVIATIONS AND TERMS
AI Artificial intelligence
HPL The Private Employment Agencies Association in Fin- land
HR Human Resources
RPO Recruitment process outsourcing
1 INTRODUCTION
1.1 Research background
As societies develop, the rapidly evolving technologies offer new opportunities.
Companies want to keep up with the newest trends to develop their business within the possibilities offered by technology. One of the biggest concepts in to- day’s world is artificial intelligence. Artificial intelligence has already a visible role in the business world and everyday life. The utilization of artificial intelligence varies between companies and industries: organizations that are well advanced in digitalisation are also involved in the application of artificial intelligence. Data has become the most valuable resource in the world. (Ministry of Economic Af- fairs and Employment 2019.)
Digitalization in human resource management began in the early 1990s when the first electronic databases were created. Nowadays automation and artificial intel- ligence are common in many companies’ HR functions. (Haka 2018.) Artificial intelligence has also become a valuable tool in recruitment, and repetitive tasks are being processed by features of AI. Artificial intelligence is continuously devel- oping, and new ways of utilization are being invented in a fast pace. Algorithms, automation and machine learning are features of AI that are being used to handle aftercare, communication, creating job advertisements, and much more. It is yet uncertain what artificial intelligence will be capable of in the future, but many pre- dictions have been made on the subject matter. (Pihlajamaa 2019.) The research topic originated from the relevance of AI as it is already utilized in the recruitment process, and the role of it in recruitment is only increasing.
1.2 Thesis objective, research questions and limitations
Every research has a purpose or a goal that guides the strategic choices made during the research. The purpose of the research can be described as mapping, explanatory, descriptive, or predictive. A research can contain more than one purpose and it can evolve as the research progresses. (Hirsjärvi, Remes & Saja- vaara 2009, 137–139.)
The purpose of the research was to map out the current use and role of AI in a recruitment process and predict how the use of it will evolve in the future. The value of AI was also investigated in the research. The empirical part of the re- search was focused on the use of artificial intelligence in Finnish personnel ser- vice companies’ recruitment process.
The main research question is:
What is the current use of AI in recruitment and how the use of it will evolve in the future?
A sub research question is:
What value does AI bring to recruitment?
The research questions were selected to get a clear image on the current use of artificial intelligence in Finnish personnel service companies’ recruitment pro- cess, and to discover how the use of it will evolve in the future. The sub research question was selected to understand what kind of value artificial intelligence brings to recruitment. Recruitment is already changing from what is used to be, and it will continue to develop with the help of AI. The use of AI allows for exam- ple, applicant anonymity in the recruitment process. This allows all candidates to be equal in the recruitment process regardless of the gender or nationality, for example. (Niinistö 2020.)
Once a research topic has been selected, it must be delimited. Only after delim- iting the topic and discovering what to prove in the research one should start the collection of the references. (Hirsjärvi ym. 2009, 81.) The research was limited to Finnish personnel service companies that are a part of the Private Employment Agencies Association. These companies are focused specifically on recruiting and potentially exploiting more features of AI than companies that carry out their own recruitment. The research was limited to Private Employment Agencies As- sociation’s members to ensure the reliability of the research (Private Employment
Agencies Association 2020a). In addition, the research was limited to AI in a re- cruitment process because AI is already used in recruitment, but there is insuffi- cient information about the extent of its use.
The Private Employment Agencies Association is an industry association and the employer of private employment service companies. HPL’s aim is to develop and improve the human resource sector. The association is formed by 330 profes- sional and responsible companies in the field. The member companies provide versatile service such as, hr-, recruitment- and hiring services, employment- and change security services, and coaching and training services. (The Private Em- ployment Agencies Association 2020a.)
1.3 Theoretical framework
The thesis is based on two main concepts and theory about value creation pro- cess. The main concepts are recruitment process outsourcing and artificial intel- ligence. Recruitment process outsourcing will be covered first, and the recruit- ment process is introduced by presenting a personnel service company Adecco’s recruitment process. After recruitment process outsourcing, AI is introduced as a concept, and the current use of it in recruitment is described. The challenges and the future of AI in recruitment will be covered as well.
A company must understand the value creation process, define it specifically and analyse the customer’s concerns to be competitive. Value creation process is built from three elements: defining the value that the company can provide for its customers, defining what value the company receives from its customers, and the exchange of these values. (Payne, n.d., 2.) Value creation is an essential part in creating long-lasting customer relationships.
1.4 Research methodology and data collection
The empirical part of the research was carried out by a quantitative research with qualitative features. An online survey was sent to the member companies of the Private Employment Agencies Association in Finland. The survey was sent by email and the contacts for the personnel service companies were collected from
the Private Employment Agencies Association’s website. The data were collected and analysed by using a survey tool called SurveyPal.
Figure 1 supports the research schedule. The thesis process started in the be- ginning of August 2020 with the selection of the thesis topic. Once the topic was selected, a research plan was constructed. During September 2020 the re- searcher formed much of the theoretical framework of the thesis. The theoretical framework was created first to ensure that the researcher had enough knowledge about the topic before crafting the survey. In the beginning of October 2020, the researcher started to craft the survey and collect the respondent emails. The sur- vey was sent out on the 14th of October 2020 for the respondent companies and companies were given time until the 23rd of October 2020 to answer the survey.
During this time the researcher supplemented the theoretical framework. A re- minder email of the survey was sent on the 20th of October 2020. On the 26th of October 2020, the researcher started to go through the responses and formed the results part of the research. In the beginning of November 2020, the re- searcher started to analyse the received data. After finishing the analysation of the data, the researcher wrote the conclusion and abstract, and finalized the the- sis by the 20th of November 2020.
FIGURE 1. Thesis timeline
1.5 Thesis structure
Figure 2 below demonstrates the structure of the thesis. The thesis will start with an introduction of the topic which explains the research background, objective, research questions, limitations, theoretical framework, and research methodol- ogy and data collection. After this theory about recruitment process outsourcing
and AI in recruitment are introduced. After the theoretical framework, the empiri- cal analysis is presented. Finally, the thesis will end with a conclusion of the re- sults and the research.
FIGURE 2. Thesis Structure
2 RECRUITMENT PROCESS OUTSOURCING
This chapter will define recruitment process outsourcing as a concept and explain the main reasons for outsourcing. A recruitment consulting process in a person- nel service company Adecco will be introduced.
Recruitment can be defined as the process of finding the best candidates for a position and eventually hiring a person. Recruitment can be done for an already vacant job or an anticipated opening. The goal of recruitment is to form an ap- pealing job advertisement and to provide enough information about the job de- scription and specification to find potential applicants. (Kazi 2018.)
Outsourcing can be defined as an action where an internal function of the organ- ization is transferred to an external service provider (Rajala, Tammi & Meklin 2008, 45–46). Recruitment process outsourcing means the process of transfer- ring all or a part of a company’s recruitment to an external provider (Resource solutions 2020). The purpose of recruitment process outsourcing is to bring value to the customer companies by making the recruitment process as easy and ef- fortless as possible, and by finding the most suitable employee for the job de- scription and the work community.
In the recent years competition and digitalisation have contributed into the growth of service outsourcing which has affected the outsourcing of recruitment as well.
Recruitment is a time-consuming process and a long-time investment, and many organizations lack the resources for doing it properly. (Studentwork 2019.)
There are many reasons for recruitment process outsourcing. Companies and organizations want to avoid making a poor recruitment decision as it can become expensive. Personnel service companies have professionals handling the recruit- ment and a wide network to reach the potential employees. Therefore, the re- cruitment process is often faster and the possibility of finding a suitable employee is higher. Securing good applicants by being fast and efficient are key factors in the process. Recruiting by using a personnel service company also allows the company to have an objective recruitment and time to put emphasis on the com- pany’s core business and operations. (Studentwork 2019.)
2.1 Recruitment process outsourcing
A senior recruitment consult Mira Perttula was interviewed in 2018 as part of a research about personnel service company Adecco’s recruitment process. Next Adecco’s recruitment consulting process will be introduced in detail. Figure 4 summarizes the recruitment steps of Adecco.
2.1.1 Assignment and initiating the recruitment
The recruitment process starts with a customer need and an assignment from the customer company. An offer is made to the customer company and once it has been approved the next step is a meeting with the customer company. In the meeting a schedule for the recruitment is discussed and a job analysis is defined.
The customer company describes the open position and the criteria that needs to be fulfilled. For example, job specification; the amount of experience, certain education, or characteristics. These details are defined carefully to facilitate the process of finding the right person for the job. Job description is also agreed on which elaborates the tasks included in the job, why and how the job is done and what are the goals for these tasks (Kazi 2018). Setting clear and defined criteria guides the recruitment process. With a clear criteria it is easier to understand what the company is looking for in an employee. (Salli & Takatalo 2014, 15.)
A job advertisement is prepared and approved by the customer company. After the approval it is published in the personnel service company’s website and in other job search channels. The application period is approximately two weeks.
(Cugal & Rankila 2018.) When choosing the channels, it is important to keep in mind the target group. The recruitment channels must be chosen based on the target group’ preferences and where they are active to reach the most potential applicants.
A job advertisement should be appealing and succinct to attract the applicants.
The open position should be marketed from the applicant’s perspective; why the applicant should apply to this exact position and company. The open position should be introduced with a performance-based description to give the applicant a concreate image on what the tasks include. The job advertisement should also
be marketed through the channels that the potential applicants are using. The advertisement can and sometimes should be published in multiple channels to reach all the potential applicants. (Salli & Takatalo 2014, 25–28.)
Figure 3 presents the elements of a good job advertisement. A job advertisement should be visually appealing as visual things are often more memorable and it can increase brand awareness and visibility. The open position should be easy to apply, and the instructions should be clear. The instructions should clearly state what the applicant must submit, and the application should not be too laborious.
Filling the same information multiple times can give a bad impression of the ap- plication process. (Salli & Takatalo 2014, 28–29.)
FIGURE 3. A good job advertisement (Salli & Takatalo 2014, 25).
If the need for a new employee is urgent, the personnel service company’s exist- ing database of employees is used. Already registered applicants that filled their open application and meet the criteria for the place are interviewed. (Cugal &
Rankila 2018.) A method where a suitable employee is recruited without publish- ing a job advertisement is called executive search or headhunting. Headhunting is usually used with positions for senior management or demanding specialist positions. (StaffMill 2020a.) With a big database of possible employees, the re- cruitment process can be accelerated, and the customer need is fulfilled in a fast pace.
2.1.2 Interview process and selection
Once the application period is over the initial screening can begin. A consultant goes through all the applications and invites all the suitable applicants for a pre- interview. Pre-interviews are handled by calling the applicants. Every applicant is often called unless a clear deficiency of important factors appears, such as expe- rience or education. (Cugal & Rankila 2018.)
Applicants that seem to fit the job description the best proceed to the second round of interviews that are held face-to-face. If the applicant is far away or for some reason cannot attend the interview physically then a video interview is held.
With an urgent need, a face-to-face interview held by the consultant is omitted and the applicant will proceed to an interview held by the customer company.
(Cugal & Rankila 2018.)
Based on the overall interview process two to four candidates that are well-suited for the position are introduced for the customer company. If one of the suggested applicants is chosen for the position, the personnel service company will inform the candidate. Rejected candidates are also informed about the decision. (Cugal
& Rankila 2018.) Applicants that are not chosen for the position will more likely apply again for a position in the same company if the aftercare is taken care of.
2.1.3 Interview questions and methods
The questions asked during the interview process are based on the customer company’s requirements and the job specification. The interview questions are conducted by the recruitment consultant. The consultants use both structured and content interview styles and the combination of them. The consultant can use
situational interview questions to find out how the applicant would behave in cer- tain job-related situations. Behavioral interview questions can be also used to detect how the applicant has acted in past circumstances that may be linked to the position they are applying to. This can be for example, asking about a difficult customer service situation (Salli & Takatalo 2014, 62). Stress interview methods are also used sometimes to see how the applicant reacts to stressful and sudden situations. This can be for example, a sudden change of the interview language.
(Cugal & Rankila 2018.)
2.1.4 Selection, billing and aftercare
If the customer company employs an applicant through the personnel service company, the selection process will end with billing. The length of the recruitment process from the job analysis until the selection is approximately four weeks.
When the customer company has an urgent need, the length of the process is shorter. Also, with long-term customer relationships the recruitment process is often shorter as the personnel service company is already familiar with the com- pany and possible need. (Cugal & Rankila 2018.)
After a successful recruitment, the customer company and the chosen employee are contacted once the employee has already started in the position. This is to ensure that the employee has adjusted into the work environment and to check how the employee has been performing in the position. (Cugal & Rankila 2018.) The aftercare is an important part of the recruitment process as the personnel service company’s value is increased from the customer company and the appli- cants’ point of view.
FIGURE 4. Adecco’s Recruitment Steps
3 ARTIFICIAL INTELLIGENCE IN RECRUITMENT
The progress of technology has been rapid for the past decade and one concept has been rising above others. This concept is artificial intelligence, in short AI.
This chapter will define artificial intelligence as a concept and the features it pos- sesses. This section will address the current use of artificial intelligence in a re- cruitment process and how the utilization of it will develop in the future.
3.1 Artificial intelligence
Artificial intelligence is a subfield of computer science. AI has been defined in multiple different ways for as long as it has existed. The definition is continuously changing as the elements and features of AI are developed.
As the term artificial intelligence is difficult to define, it is sometimes suggested to rather determine the features that it possesses. The main features of AI are au- tonomy and adaptivity. Autonomy is defined as an ability to operate in a complex environment without constant user control. Adaptivity means the ability to en- hance performance by learning from previous experiences. (University of Helsinki
& Reaktor 2020a.)
Artificial refers to something that is simulated and does not exist naturally. Intelli- gence means the ability to learn, think and understand. The methods in creating AI are using data and different algorithms but the main concept is that AI eventu- ally starts to learn to create itself and improve the algorithms. (PwC 2017.)
Artificial intelligence can be divided into strong and weak AI. Strong AI is defined as a being that is truly intellectual and conscious. The aim of strong AI is to de- velop the intellectual capability of machines to function as well as a human being.
Reportedly strong AI has not been developed successfully yet. Weak AI refers to the current AI systems that are capable of implementing intelligent functions.
(University of Helsinki & Reaktor 2020b.) Artificial intelligence is capable of per- forming tasks that simulate intelligent human activity such as, inference, learning, anticipation, decision-making, and speech and character recognition. (Aureolis 2020).
Machine learning is one subfield of artificial intelligence and most functions of AI are based on it. Machine learning is described as the systems that enhance their performance in a particular task as more data and experience are accumulated.
This means that there are no defined operating instructions for each possible scenario, but the machine learns from the teaching data independently. The goal of machine learning is to automate the interpretation of information and expand the machine's ability to make observations. Machine learning enables AI applica- tions to be adaptive. Machine learning can be divided into supervised, unsuper- vised, and reinforcement learning. (University of Helsinki & Reaktor 2020c.)
Machine learning is developed to support decision-making, customer service and streamlining production processes. To take advantage of machine learning it is provided that the business problem is clearly defined, the machine learning mod- els are integrated properly as part of the company’s operations, and that the used data is high quality. (Aureolis 2020.)
Robotics is also a part of artificial intelligence. It involves building and program- ming devices that can operate in a complex real world. Robotics requires almost all the sub-areas of AI and therefore is the most intricate part of AI. (University of Helsinki & Reaktor 2020c.)
The process of implementing artificial intelligence as part of a company’s opera- tions starts with business problem and data mapping. After that AI is tested on a trial material and the business profit of it is evaluated. (Aureolis 2020.) With a successful implementation of artificial intelligence, the company earns value to their customers, and therefore to the company’s business as well.
3.2 Utilization of AI in recruitment
There is already evidence of the utilization of artificial intelligence in recruitment.
AI can be used in many parts of the recruitment to support a more effective pro- cess. In this chapter the current utilization of AI in a recruitment process is intro- duced.
3.2.1 Job application processing and matching
Personnel service company Barona has developed their own recruitment envi- ronment called Jelpp. Jelpp was developed to assist the recruiter in daily tasks.
The AI used in Jelpp pre-screens applicants, matches them with positions based on the applications, and searches the application database for suitable appli- cants. By going through multiple job applications every week, and suggesting applicants for suitable open positions, Jelpp expedites the recruitment process for both, the applicants, and the recruiters. The process of improving and devel- oping Jelpp is continuous. (Talouselämä 2018.)
TalentAdore is another recruitment tool supported by the features of AI. It is a virtual recruitment assistant handling many parts of the recruitment process, such as matching and personal applicant communication. The applicant matching rec- ommends applicants to positions that they seem to be applicable for. The appli- cant matching uses a context search to screen and arrange the applicants based on suitability. The context search ensures that the applicants are not recom- mended a position only based on the right key words but based on their holistic background. The virtual recruitment assistant uses traffic lights to express the suitability of the applicants. (TalentAdore 2020.)
3.2.2 Applicant communication and customer service
Modern recruitment process requires active and relevant applicant communica- tion. The given information should be personal and timed correctly. (Duunitori 2018.) Job applicants want to know updates about their application, interviews, possible acceptance or rejection and AI can play a great role in it.
According to Duunitori (2018) applicant communication can be divided into four critical stages. The reception of the application, invitation to an interview or rejec- tion, information about the schedules, and the final decision. When an applicant is correctly and actively informed, they are more invested in the recruitment pro- cess. Even if they are not selected for the position, they are more likely to have a positive brand image of the company and recommend it to others. Companies nowadays use CRM and recruitment systems that are guiding the communication
to a better direction. With the help of AI and the automatization of the messages it also intensifies the recruiter’s time management with other important tasks.
TalentAdore’s virtual recruitment assistant was created to enhance also the ap- plicant communication among other tasks. The applicant communication is per- sonalized by using the data about the applicant. The features of AI transform the information from the recruiter’s clicks into a natural language and to a personal- ized message for the applicant. TalentAdore’s applicant communication is used to give feedback, status updates and for arranging job interviews. (TalentAdore 2020.)
Customer service is an essential part of the company’s functions. To attract po- tential individuals to apply for the open position the customer service must be in order and the response speed fast. Company’s brand image is built for the most part on customer experiences and fast customer service is one crucial part in it.
This is where the utilization of chatbots is beneficial.
There are many types of chatbots but according to Ekman (2019) chatbot’s three main utilities are that it asks, answers and screens automatically. The questions and answers provide valuable data and ideally chatbots can pre-screen the most potential applicants among many. Especially with large customer pools chatbots are useful as they can pass smaller amount of higher quality applicants for the recruiter to review. Chatbot can also give information about the open positions that can motivate the individual to apply instantaneously. Chatbots can save the time of recruiters’ and work as a vital tool in creating a great applicant experience.
3.2.3 Job advertisement keyword planning and targeting
When creating a job advertisement, a company is creating its first impression for potential applicants. A good job advertisement can build a good brand image and in order to achieve that the employer must consider what applicants find appeal- ing. (HR Viesti 2019.) Keyword planning is used to detect attractive word choices to use in job descriptions and choices that reach even the best candidates.
Textio is an AI-powered tool that offers help with writing an appealing job descrip- tion. Textio gives suggestions for appealing word choices, phrases, sentences, and even paragraphs. It also gives improvement ideas. It can for example, detect pressuring tone and recommend avoiding certain words or sentences. The sug- gestions are based on an extensive and up-to-date data on language perfor- mance. Textio also shows how companies’ hiring language is performing among their competition. (Textio 2020.)
Job advertisements targeting is a smart way of exploiting AI. With the use of al- gorithms, ads can be targeted to the right target group and to candidates that have recently viewed similar job advertisements. With the help of AI, job adver- tisements can be shown in the potential applicant’s social media or application portal. (Pihlajamaa 2019.)
3.2.4 Interviewing
Artificial intelligence can now function as an interviewer as well. Job interviewing robot, Robot Vera, can interview multiple candidates simultaneously and analyse their suitability based on their body language and micro impressions. (Pihlajamaa 2019.) Robot Vera receives detailed job description, list of skills, and experience required, and interview questions before beginning the process of finding suitable candidates. After the approval of a recruiter, Robot Vera can schedule the inter- views and call them on the agreed time. Robot Vera uses a voice-recognition and a vocabulary of 13 000 words to interact with the candidates naturally. (J.B.F.
News 2018.)
3.3 Challenges of AI in recruitment
Artificial intelligence-related risks and threats are often affiliated with the rise of machines and the destruction of humanity. According to Helsinki University’s as- sistant professor Teemu Roos, these concerns are unnecessary, but there are risks in the use of artificial intelligence. (Kemppi 2018.)
One of the biggest challenges of AI in the recruitment process is caused by ma- chine learning and algorithmic bias. Algorithmic bias means the machine’s une- qual decision making based on the applicant’s gender, age, or ethnicity in the recruitment process. Algorithmic bias is caused by discrimination by people. Ar- tificial intelligence is as smart and reliable as the data that it utilizes. For example, if the machine has received data where male candidates are chosen over equally good female candidates, the machine can learn to discriminate females. (Univer- sity of Helsinki & Reaktor 2020d.) A case of algorithmic bias happened with Am- azon’s recruiting machine. The system was not rating the candidates in a gender- neutral way and the machine learned to favour men based on the data that was inputted into it. (Pihlajamaa 2019.)
Another issue about AI was raised by an American researcher Dan Greene in Yle Tiede Work2017 conference. He mentions that there is a concern that the use of AI in a recruitment process develops a group of people that will be marginalized outside the labour market. The final decision is always based on a face-to-face interview, but the problem is that some candidates might never make it there.
Because artificial intelligence learns efficiently the attitudes and prejudices from the inputted data, it can cause unfair selection in the recruitment processes. (Nur- milaakso 2017.) And despite all the data available in the world, machines and AI still lack the humanity of people. No machine will be able to fully capture this feature and therefore AI will remain as a supporting tool for humans. (Pihlajamaa 2019.)
3.4 Future of AI in recruitment
During this millennium the development of AI has been rising. Breaking down big problems into smaller sub-problems has affected the progress. Currently AI has been focused on solving practical and significant real-world problems instead of large entities that attempt to solve general artificial intelligence that solves all problems. (Kemppi 2018; University of Helsinki & Reaktor 2020.)
As a topic artificial intelligence provokes a lot of discussion, uncertainty and even fear. Even if AI still possesses challenges many positive and progressing things
can happen in the future of recruitment with the help of artificial intelligence. Ac- cording to Duunitori’s national recruitment survey (2020), the direction of the re- cruitment industry is going towards a candidate-driven recruitment and digitaliza- tion will be an essential part in it. The application process will be shaped to be clearer and more personal, with the help of chatbots, video interviews and more efficient applicant communication. Electronic channels will be the main place for recruitments.
According to the Barona Technologies chief technology officer Kirsi Louhelainen, Barona seeks constantly new opportunities for the use of AI. Their recruitment environment Jelpp will be developed even further to assist the job seekers, and they see AI as a digital assistant that brings value to their company and custom- ers. However, Louhelainen mentions that AI will not replace the recruiter or the human reasoning. They are using Jelpp as a support for the recruiter and not as an individual tool. (Talouselämä 2018.) The Bolt Works CEO Tommi Hämäläinen says that he sees artificial intelligence as the future of recruitment. With the help of AI, encounters between employers and employees can be improved. AI has a positive impact on recruitment, but also Hämäläinen agrees that AI alone is not enough. (Kalska 2018.)
The labour market also in the public sector will evolve in the near future. Job market Finland is a new upcoming job market service developed by KEHA-Centre and the Ministry of Economic Affairs and Employment. It provides a shared plat- form for public, private, and third sector providers of employment and training services, and for various employment exchange platforms. With the help of Job market Finland, job applicants are offered better possibilities for finding work that corresponds to their own skills and knowledge. In addition, employers will obtain better tools for finding new, suitable employees. The newest technologies and artificial intelligence will be utilized in the process. (KEHA-Centre 2020.)
According to LinkedIn’s talent blog, recruiting will be changing already within five years. Recruiting will have even bigger and more important role because talent will matter more. Because automation and AI is replacing the need for individuals to do repetitive tasks, the work that is left for humans require more creativeness.
Hiring needs will evolve as business models and markets change and more re- cruitment tools and technology will be invented and developed. (Lewis 2019.)
In the future recruitment processes will be more automated, AI will be utilized more, job descriptions will evolve, and the process will be faster. The forerunners will be the companies that are able to manage and organize their own workflow the best. (StaffMill 2020b.) But even if AI will be a significant part of the future in recruitment, recruitment is a process that will always require human contact even at some level. Artificial intelligence can support and facilitate the recruitment pro- cess, but the final decision will require human judgment and decision making also in the future. (Pihlajamaa 2019.)
4 RESULTS AND DATA ANALYSIS
The aim of the empirical part of the research was to gather more information about how artificial intelligence is utilized in the recruitment process currently and how the use of it will progress in the future. In this chapter the research methods and the results of the research are introduced and analysed. The questions of the survey can be found in appendices (appendix 1).
4.1 Research methods and data collection
The empirical part of the thesis was executed by conducting a quantitative re- search with qualitative features. According to Vilkka (2007, 26) quantitative re- search seeks to provide a general description of the results numerically. The aim of the quantitative research is to discover general principals. A research can have both quantitative and qualitative research methods, and they can be used as sup- plementary orientations (Hirsjärvi ym. 2009, 136–137). This research, while being a quantitative research, contains qualitative features. The open questions are an- alysed in a qualitative way. Qualitative research seeks to represent real life, and humans are preferred as the instrument of knowledge. (Hirsjärvi ym. 2009, 164.)
An online survey was selected as the quantitative research method to obtain ex- tensive research data. A survey is a research collection method in which all re- spondents are asked the same questions, in the same order and in the same way. The format of the questions is standardized. Survey is used when the re- spondent’s attitude, opinions or behaviour is being researched. (Vilkka 2007, 27–
28.)
The aim of the survey was to attain wide range of responds from different com- panies and across Finland. Given the resources and the extensive pool of exam- inees, an online survey proved to be the most practical and effective option for the research. An online survey as a data collection method is common nowadays and the geographical accessibility is easier. In addition, online survey allowed respondents to participate in the survey according to their own schedules.
The survey about artificial intelligence in recruitment was sent to 150 different Finnish personnel service companies across Finland. The contact information was gathered from the Private Employment Agencies Association’s website. The survey was sent in total to 150 member companies out of the total 330. Many of the member companies in HPL’s website were subsidiaries and therefore the re- searcher used sampling to ensure that one organization can only participate once. Out of the 150 email addresses, three company’s email addresses were invalid.
An online survey about the use of artificial intelligence in recruitment was sent to Finnish personnel service companies that were members of the Private Employ- ment Agencies Association in Finland (appendix 1). The survey was made with a tool called SurveyPal and sent by email on the 14th of October 2020 with a cover letter (appendix 2). The respondents were given time between 14th and 23rd of October 2020 to participate. A reminder email of the survey was sent on Tuesday 20th of October 2020 (appendix 3). A reminder of the survey was sent to improve the response rate of the research. The survey was tested before it was sent to the respondent companies to ensure the functionality of it.
The question types used in the survey were multiple choice, open-ended and scale-based questions. These question types were used to obtain as much infor- mation as possible from the respondents. A combination of a structured question and an open-ended question was also used to gain perspectives that the re- searcher had not been able to reason in advance. (Hirsjärvi ym. 2009, 198–199.)
4.2 Results
The online survey conducted for the companies was executed in Finnish due to the difficult nature of the topic, and because the companies are from Finland. The results are introduced in English. Results of the survey will be sent to companies that were invited to participate in the survey. The survey was sent to 150 different personnel service companies. The total respondent number was 40. The re- sponse rate was approximately 27 %.
4.2.1 Respondent background
Respondents were asked four background questions. These questions were about gender, age, professional title, and recruiting history. The survey started with the background questions because the topic of the survey was difficult.
As can be seen from figure 5 majority of the respondents were females. Out of the 40 respondents, 58 % of the respondents were female and 43 % male.
FIGURE 5. Respondent gender
Figure 6 demonstrates the age distribution. Out of the 40 respondents, only 3 % was under 25 y/o, 33 % was 25-35 y/o, 30 % were 36-45 y/o, and the rest 35 % were over 45 y/o. Most of the respondents were over 45 y/o.
FIGURE 6. Respondent age
Out of the 40 respondents, 39 responded to the question about their professional title. Figure 7 shows the distribution in the professional titles. Out of the 39 re- spondents, 36 % were working in the executive team and 39 % as a recruitment consultant or a recruitment coordinator. The rest 26 % answered that their pro- fessional title is something else, such as an entrepreneur, recruitment manager, researcher, sales manager, or hr-specialist.
FIGURE 7. Respondent professional title
The last of the background questions was about the respondent recruiting history.
Out of the 40 respondents, 39 answered to the question about their recruiting history. Shown in figure 8 is that most of the respondents have been working in recruitment for over 5 years. Out of the 39 respondents, 15 % have been working in recruiting less than 2 years, 33 % have been working in recruiting for 2-5 years, and 51 % have been working in recruiting for over 5 years.
FIGURE 8. Respondent recruiting history
4.2.2 What kind of image do you have about artificial intelligence in gen- eral?
Figure 9 demonstrates the respondents’ image about artificial intelligence in gen- eral. Out of the 40 respondents, 39 answered to the question about it. The re- spondents were asked to answer from a scale of 1-5, very negative to very posi- tive. The average was 3,51, between options neutral and quite positive. Majority, 53,8 %, of the respondents felt neutral about artificial intelligence in general and they do not have necessarily a negative or positive image about it. None of the respondents stated that they have a very negative image about AI in general, and only 5,1 % said that they have a quite negative image about it. 25,6 % of the respondents stated that they have a quite positive about image about AI in gen- eral, and 15,4 % of the respondents said to have a very positive image about it.
FIGURE 9. What kind of image do you have about artificial intelligence in gen- eral?
4.2.3 Is artificial intelligence currently being utilized in your company’s recruitment process?
All respondents answered either yes or no when asked if artificial intelligence is currently utilized in their company’s recruitment process according to figure 10. A vast majority of 77,5 %, replied that they are not currently using AI at all in their company’s recruitment process. The remaining 22,5 % stated that they are cur- rently using AI in their company’s recruitment process.
If the respondent answered No or I don’t know, they did not have to answer to question number 4, about the current use of AI in their company. If the respondent answered yes, they did not have to answer to question number 3, about their company’s possible plans of introducing AI to their recruitment.
FIGURE 10. Is artificial intelligence currently being utilized in your company’s re- cruitment process?
4.2.4 Does your company have any plans to introduce artificial intelli- gence in the recruitment process in the future?
Respondents were asked, according to figure 11, their company’s plans in the future with AI. Out of 40 respondents, 31 replied in the previous question that they do not utilize AI answered to the question. Out of those 31 respondents, 41,9
% did not know their company’s plans with AI, 19,4 % replied that they have plans to utilize AI in the future, and 38,7 % replied that they have no plans to introduce AI in their company’s recruitment process.
FIGURE 11. Does your company have any plans to introduce artificial intelligence in the future?
4.2.5 In which areas of recruitment is artificial intelligence currently uti- lized in your company?
The current use of artificial intelligence in the respondent companies was mapped out by using a multiple-choice question with an open-ended question on how AI is being utilized. Multiple-choice question allowed the respondents to select more than one option. An open-ended question was used to gain concrete examples on the use of AI. All of the respondents did not answer to the open-question part.
Figure 12 demonstrates the current use of AI in those companies that said that they are already using AI in their recruitment process.
From the nine (9) respondents that said to utilize AI in their company, three (3) answered that they utilize it in customer service. The use of artificial intelligence in customer service were the use of a chatbot and a phone bot. One (1) respond- ent answered that they use AI in interview arranging and scheduling by using
automated calendar entries from their own system. Five (5) respondents men- tioned that they use AI in application reviewing and preselection. Four (4) out of the five (5) respondents mention that the utilization occurs as the use of an appli- cant tracking platform that suggests the best suited applicants for the wanted position. One respondent remarks that the accuracy of the system has been pre- cise. Another respondent on the contrary mentions that the system they used is not in operation anymore due to the poor functionality of it. In addition, another respondent mentions that the weight for the platform’s suggestion is still low and all applications are processed manually and equally. One (1) respondent an- swered that they use a phone bot in the application reviewing and preselection.
When asked about the use of AI in arranging interviews or in the interview, none of the respondents mentioned that they use AI for those purposes. Four (4) re- spondents said that they use AI in applicant communication. Three (3) of those respondents mentioned that they used automated emails based on the applicant status. In addition, four (4) respondents mentioned that they use AI in something else during their recruitment process. One (1) respondent mentioned that they use AI in applicant evaluation and another respondent in identifying and finding potential customers. One (1) respondent says that they use AI in research where LinkedIn brings up interesting profiles based on past search history. The weight on this function is low because the assignments vary a lot and the suggestions arrive late, often after the end of the assignment. One (1) respondent also men- tions that they are currently testing a system that gives a possibility for all the above-mentioned points, but it is not yet in active use.
FIGURE 12. In which areas of recruitment is artificial intelligence currently utilized in your company?
4.2.6 Do you feel that artificial intelligence has a positive impact on re- cruitment?
Next respondents were asked how positively they felt that artificial intelligence has impacted recruiting. Out of the 40 respondents, 39 answered to the question about it. The respondents were asked to answer from a scale of 1-5, not at all positive to very positive. The average response was 3,41.
Shown in figure 13, 43,6 % of the respondents answered neutral when asked whether they feel that AI has a positive impact on recruitment. None of the re- spondents answered not at all positive and 10,3 % answered not very positive.
Out of the 39 respondents, 41 % felt that AI has impacted recruitment quite pos- itively and 5,1 % felt that the impact has been very positive.
FIGURE 13. Do you feel that artificial intelligence has a positive impact on re- cruitment?
In this question respondents were asked to briefly explain their answer. Many of the respondents who answered that AI has quite positive or very positive impact on recruitment mentioned that with the help of AI there will be more time left for the most important parts of recruitment. The repetitive and routine tasks can be supported by AI and automation, to decrease the human errors, fasten the pro- cess, and to increase the time for the company’s core business. Some respond- ents also mentioned that with the use of AI, the applicant and open position can be connected more efficiently, and applicants can be helped to better express their skills. One respondent mentioned that especially in the recruitment of soft- ware development-type tasks, AI would streamline the recruitment process, be- cause know-how could be measured by various competence-based tests. Some respondents also mentioned that even though AI might be useful in some parts of the recruitment, the selection of the applicants is impossible without a human’s decision-making.
Many of the respondents who answered neutral when asked if AI has had a pos- itive impact on recruitment, mentioned that currently the use of AI is still quite small in recruitment that they had not seen the positive effects yet. Many respond- ents also pointed out that even if the routine tasks can be replaced with AI, it is still not competent enough to make decisions about the more important and diffi- cult parts of recruitment. Some of the respondents had also used AI in their com- pany and the functionality of it had not been successful. Few respondents also mentioned that they do not have enough knowledge about the topic to have an opinion, but most likely in the future AI will be more involved in recruitment.
Two respondents who answered not very positive when asked if AI has had a positive impact on recruitment justified their answer. The other respondent men- tioned that based on their experience with AI they do not believe that it is func- tional in recruitment. They also added that if AI could be properly integrated it could be beneficial mostly in the beginning of the recruitment process. The other respondent said that in their recruitment process they require a meeting with the applicants because of their fast pace in recruitment for substitute places, for ex- ample.
4.2.7 Do you feel that artificial intelligence has changed recruitment?
Respondents were asked if they felt that artificial intelligence has changed re- cruitment, according to figure 14. Out of the 40 respondents, 36 answered to the question about it. The respondents were asked to answer from a scale of 1-5, not at all to quite a lot. The average was 2,86.
Half of the respondents (50 %) answered neutral when asked if artificial intelli- gence has changed recruitment, 25 % answered not much, and 5,6 % answered not at all. Out of the respondents 16,7 % answered that AI has changed recruit- ment quite a lot and 2,8 % answered that it has changed it very much.
FIGURE 14. Do you feel that artificial intelligence has changed recruitment?
In this question respondents were asked to briefly explain their answer. Some respondents who answered quite a lot or very much when asked if AI has changed recruitment mentioned that AI has fastened the process. One respond- ent also mentioned that it has also made the separation of individuals harder.
Another respondent said that partly it has made the recruitment process more impersonal. In addition, one respondent pointed out that in recruitment it is im- portant to be aware of the current trends and find the competitive advantages, AI being one of them.
Majority of the respondents who answered neutral when asked if AI has changed recruitment mentioned that the progress and usability of AI is not yet far enough and therefore, it has not changed recruitment yet. Some respondents mentioned that the process might be a bit more automated nowadays, but the basic elements of the recruitment process has not changed. In addition, some of the respondents stated that they do not have enough knowledge about the topic and therefore they responded neutral.
Majority of the respondents who answered not much or not at all when asked if AI has changed recruitment, said that AI is not yet widely utilized in recruitment and therefore, has not changed recruitment. However, many of the respondents mentioned that probably AI will change recruitment, but in the future. Some re- spondents also mentioned that AI is still not ready for parts of recruitment, such as in-depth interviews.
4.2.8 Do you feel that artificial intelligence is going to change recruiting in the future?
The seventh question concerned about the future of artificial intelligence, and the respondents were asked if they felt that AI will change recruiting in the future according to figure 15. Out of the 40 respondents, 37 answered to the question about it. The respondents were asked to answer from a scale of 1-5, not at all to quite a lot. The average was 3,95.
Many of the respondents felt that recruiting will be changing quite a lot in the future. Out of the respondents, 59,5 % answered that recruiting will change in the future quite a lot, 18,9 % of the respondents felt neutral about it and 18,9 % felt that it will change very much in the future. None of the respondents felt that AI will not be changing recruitment at all, and only 2,7 % felt that it will not change it much.
FIGURE 15. Do you feel that artificial intelligence is going to change recruiting in the future?
In this question respondents were asked to briefly explain their answer. Majority of the respondents who answered that AI is going to change recruitment in the future quite a lot or very much mentioned that AI will develop and therefore it will be more involved in recruitment as well. The automation and replacing of certain tasks with AI will change the recruitment process. Some respondents mentioned that this might affect positively in terms of the speed of the process, and prioritiz- ing of the most important tasks. On the contrary, it might also affect negatively if the applicants feel that they are only interacting with a chatbot, for example. Many of the respondents also mentioned that AI will be used in the future in more de- manding tasks and the use of AI can decrease prejudices, but AI will not replace humans’ value in recruitment. In addition, one respondent said that they believe that their own “productive” position will be completely replaced by artificial intelli- gence applications in the future.
Respondents who answered neutral when asked if AI will change recruiting in the future, mainly felt that it might change a bit but not significantly. One respondent mentioned that it can be used in the routine tasks, but it is not advanced enough to make judgements on today’s requirements. One respondent who answered that AI will not change recruiting a lot in the future said that the change will be a positive one, by supporting the repetitive and routine tasks.
4.2.9 In which areas of recruitment do you find artificial intelligence to be the most useful?
Figure 16 demonstrates the respondent’s opinions on the most useful areas of recruitment to be supported by artificial intelligence. The areas were mapped out by using a multiple-choice question. Multiple-choice question allowed the re- spondents to select more than one option. Out of the total 40 respondents, 39 answered to the question.
Nine (9) respondents stated that AI would be useful in customer service. 26 re- spondents mentioned that it would be useful in application screening and prese- lection, making it the most selected area for AI’s utilization. 16 respondents said that AI would be useful in interview arranging and scheduling, three (3) responded the interviews, and 24 found it to be the most useful in applicant communication.
In addition to that, six (6) respondents mentioned some other area that would benefit from AI. These areas were analysing and collecting data, applicant clas- sification, sales, competence testing, matching of the applicants and the posi- tions, and in general facilitating the application process.
FIGURE 16. In which areas of recruitment do you find artificial intelligence to be the most useful?
4.2.10 Briefly describe how your company, your client companies and your job seekers will benefit from the effects of artificial intelligence
In this open-ended question respondents were asked to describe how their com- pany, customer companies, and job seeker will benefit from the effects of artificial intelligence. 20 respondents, out of the total 40, answered to the question about it.
Many respondents mentioned the fastening of the recruitment process as one of the benefits, which will benefit all parties involved. Many also stated that the ap- plicants will receive answers to their questions more quickly, the errors will be minimized, and the recruiters’ time can be saved. Some respondents also men- tioned that currently the benefits are quite minimal, but once the features of AI are improved and can be utilized more, the effects are more visible. Some re- spondents mentioned that they do not utilize AI currently and therefore, there are no benefits of AI.
4.2.11 Do you have anything else to add?
Majority of the respondents did not have anything to add. One respondent pointed out that the use of artificial intelligence will unfold properly once it is actually uti- lized. Another respondent added that in the recruitment industry, all work steps that do not require direct personal communication between people will be out- sourced to artificial intelligence applications eventually. This allows the resources to be directed at the company’s core competencies, which is the communication to recognise an individual’s personality, experience and competence.
4.3 Analysis of the research material
After the material has been collected it needs to be analysed, interpreted, and final conclusions must be done. In the analysis part, the researcher will carefully analyse the data that was received and concludes the results of the research.
(Hirsjärvi ym. 2009, 221.) The data was analysed with the help of SurveyPal.
SurveyPal was used to collect the answers question by question. The results of the research provide an indicative image about Finnish personnel service com- panies’ utilization of artificial intelligence in recruitment and the future of AI in recruitment.
Based on the data that was received majority of the respondents were female, over 45 years old, worked as a recruitment consultant or coordinator, and had worked in recruitment for over 5 years. The general image about artificial intelli- gence was mainly neutral, but still more positive than negative. However, the uti- lization of AI currently is low. Only nine (9) out of the 40 respondents stated that they are currently utilizing artificial intelligence in their company’s recruitment pro- cess. Some companies had plans about utilizing AI in the future, but most of the respondents did not know their companies plans with AI, or they did not have plans to utilize it.
Respondents that already utilized artificial intelligence in their company had, in general, a bit more positive image about AI, compared to respondents that did not use AI in their company. Respondents that already utilized AI in their company
mainly felt that it has had a positive impact on recruitment. In addition, those re- spondents also had wider responses and more knowledge about the topic which was expected.
The three areas that companies were mostly utilizing AI were applicant reviewing and preselection, applicant communication, and customer service. In applicant reviewing and preselection, many of the respondents mentioned that their com- pany uses a system that highlights and recommends the most suitable applicants for a certain position. In customer service respondents mentioned that they use a chatbot or a phone bot. In applicant communication the form of AI used was mostly, customized emails that are automatically edited based on the applicant’s status. Respondents felt that the use of artificial intelligence fastens the applica- tion process and brings more time for the company’s main operations.
Respondents in general mainly felt that artificial intelligence has had a positive impact on recruitment rather than negative. Some respondents did, however, mention the shortcomings of AI and those respondents felt that the best use of it would be replacing repetitive and routine tasks to save the recruiters time. Few respondents mentioned that they had tried to utilize AI in their company, but the use was not beneficial, due to unfunctional features. Many of the respondents still felt that the use of AI will increase in the future, also in the more demanding tasks.
In general, respondents did not feel that the use of artificial intelligence would have changed recruitment, because it is not widely used in recruitment yet. Vast majority of the respondents did, however, feel that AI will change recruitment in the future. Once technology and the utilization of AI increase, the errors and shortcomings of AI are repaired, many respondents felt that AI will be utilized more and the use of it will change recruitment. In the future artificial intelligence is hoped to better combine the work and the employee. Some respondents also mentioned that AI will change the job descriptions in recruitment as it can replace positions with a lot of manual work. One respondent even mentioned that they believe that their work tasks will be completely carried by AI in the future. How- ever, many respondents also stated that even if AI will change recruitment in the future it will not replace human’s empathy, decision-making, or consideration in the process.
Many of the respondents were still not too familiar with artificial intelligence which was reflected in their answers to questions where respondents were asked to answer on a scale from 1-5. Many respondents answered neutral in these ques- tions and explained that they are not enough familiar with the topic to take a stand.
In addition, some of the respondents mentioned for example, automated emails as a form of artificial intelligence. Normal automated emails are not considered as artificial intelligence anymore and as one of the respondents mentioned, the use of artificial intelligence in recruitment will unfold properly once it is being used.
Artificial intelligence as a concept is complex, broad, and quite new and therefore, it is predictable that not everyone is yet familiar with the intricacy of it.
Respondents felt that artificial intelligence would be most useful if utilized in ap- plicant screening and preselection, applicant communication, and interview ar- ranging and scheduling. The interest towards AI can be seen in this question, as all the options received quite many answers, especially the three most answered options. Over half of the respondents (26) felt that AI would be beneficial in ap- plicant screening and preselection. Out of the 40 respondents, 24 felt that it would be beneficial in applicant communication, and 16 respondents felt that AI would be useful in interview arranging and scheduling. Many of the respondents chose multiple options in this question, which indicates that the use of AI is seen as beneficial and appealing even if it is not currently being used much.
When asked about the benefits that AI brings to their company, customer com- panies, and job seekers the benefits were still quite small. Mostly respondents mentioned that the benefit for all parties involved is that the recruitment process is faster, and that the applicants can receive answers to their questions faster. In addition, decrease in the errors, and handling repetitive tasks were mentioned.
These actions save the recruiter’s time.
5 DISCUSSION AND CONCLUSION
The utilization of artificial intelligence in recruitment is truly a relevant topic. Be- cause many tasks can be replaced or supported with features of AI, it is seen as an interesting opportunity. With the help of AI, matching of the applicant and the open position happens more efficiently, and the use of chatbots as part of the customer service enhances the response time. The interest towards AI can be seen even if the utilization is not high yet.
The purpose of this thesis was to map out the current use of AI in Finnish per- sonnel service companies’ recruitment process. The thesis also examined how AI will be included in future’s recruitment and the value of AI. Based on the re- ceived answers only a small part of Finnish personnel companies is currently using artificial intelligence as part of their recruitment. It is believed that AI is not yet competent enough to handle the most important parts of recruitment such as, selection of the applicants. The biggest use of AI in Finnish personnel service companies is currently AI applications that recommend the most suitable appli- cants for a certain position. The next most popular use of AI supported parts in recruitment are customer service and applicant communication. Customer ser- vice is supported for example, by chatbots or phone bots, and in applicant com- munication customized emails that are automatically edited based on the appli- cant’s status are used. Even if the utilization is not high currently the interest to- wards AI can be seen. Companies were interested to hear the results of the re- search and in the survey the question about the most useful areas for AI in re- cruitment received many responses, and many respondents chose multiple op- tions. AI is seen as beneficial and something that adds value to the company’s operations.
Wider utilization of AI will happen in the future. Recruitment is already a more automated and digitalized process than it was before, and it is heading to this direction even more. Even respondents whose company did not use and did not have plans to use AI mainly felt that AI will change recruitment in the future. The recruitment process will be more supported by features of AI to enable more time for the company’s most important functions. AI will be used in various parts of recruitment, from time-consuming repetitive tasks to matching of the applicant
