International Journal on Artificial Intelligence Tools

Vol. XX, No. X (2025) 1–17

 World Scientific Publishing Company

The Automation of Science? Possibilities and Boundaries of AI Applications for

Conducting Systematic Literature Reviews

Przemysław Tomczyk
Department of Marketing, Kozminski University, 57/59 Jagiellonska St.,
03-301 Warsaw, Poland
[email protected]

Philipp Brüggemann
Faculty of Economics, FernUniversität in Hagen, Universitätsstraße 11
58097 Hagen, Germany
[email protected]

Tymoteusz Doligalski
SGH Warsaw School of Economics, 162 Niepodległości AV,
02-554 Warsaw, Poland
[email protected]

Received (Day Month Year)

Revised (Day Month Year)
Accepted (Day Month Year)

Science automation, driven by advanced technologies, such as artificial intelligence (AI), is

reshaping research in a variety of disciplines. The exponential growth of publications increases the
need for automation in systematic literature reviews (SLRs). In this study, we investigated how AI
applications support systematic literature reviews at different stages. After evaluating twenty-two AI
applications, we identified the most useful based on their functionalities in the SLR process. This
study provides important insights into the dynamic interplay between science automation, AI, and a
systematic literature review. The main contribution is the identification and discussion of the
relevance of the analyzed applications at different stages of the systematic literature review. We also
discuss the potential for the automation and augmentation of research activities to address various
scientific problems.

Keywords: Artificial Intelligence, Science Automation, Systematic Literature Review.

1
2 Tomczyk, Brüggemann, & Doligalski

1. Introduction
The world in which we reside is undergoing rapid transformations driven by
technological advancements and the ensuing digitalization. 1 Hence, the significance of
automation is steadily escalating and poised to emerge as a pivotal element in the process
of digital transformation2. Although automation in industries such as automobile
production has been widely embraced worldwide3, its application within the field of
scientific research remains in its infancy. According to De Bie et al.4, “given the
complexity of data science projects and related demand for human expertise, automation
has the potential to transform the data science process.” Finally, there is a pressing need
for an extensive discourse regarding the automation of scientific inquiry. The boundaries
of research automation necessitate thoughtful consideration: to what extent should
scientific work be automated and where does the optimal trade-off between facilitating
work and preserving human influence lie? Through this study, our aim was to establish a
foundation for deliberation on these inquiries by aggregating contemporary AI-driven
tools designed for systematic literature exploration. In doing so, we endeavored to
present and critically evaluate the existing landscape of capabilities, fostering a
discussion on the current state of affairs.
The expanding horizons brought forth by digitalization, encompassing various
elements such as evolving digital work paradigms, the accessibility of extensive datasets,
artificial intelligence (AI), and the emergence of large language models, are positioned to
significantly amplify scholarly research outputs in the years ahead. 1,4-6 Consequently, the
imperative role of systematic literature reviews has become increasingly pronounced in
navigating this expanding scholarly terrain. Within this dynamic context, we anticipate a
substantive potential through the integration of automation and AI. Presently, the
burgeoning volume of literature can be analyzed by employing a plethora of AI-based
tools. However, ascertaining the appropriateness of specific tools for the multifaceted
undertaking of systematic literature reviews remains a challenge. Grounded in this
reality, we extend the current research paradigm by directing our focus to the following
research question: How can AI applications bolster researchers in conducting a
systematic literature review across diverse stages?
In addressing the research query, our approach begins with an exploration of the field
of scientific automation, aiming to explicate the integration and implications of AI within
this specific domain. Subsequently, we examine the distinctive characteristics that
underlie a systematic literature review to establish the foundational comprehension
necessary for this scholarly methodology. Through exhaustive scrutiny of pertinent
scholarly works, we delineate six pivotal stages intrinsic to a systematic literature review.
Our subsequent analysis involves a critical evaluation of AI’s potential applications
across these identified stages, synthesizing insights derived from existing research
endeavors. Moving beyond this theoretical exposition, our study transitions towards the
empirical phase by explicating our research methodology and discussing the resultant
findings. This comprehensive investigation not only sheds light on the current landscape
3 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

of AI’s role in academic research but also sets the stage for potential advancements in the
domain of scientific inquiry.
Our conclusion underscores the significant potential of AI-based tools in research.
Nevertheless, their effectiveness remains contingent upon the proficiency and expertise
of researchers employing these tools. Therefore, we advocate for conscientious adoption
of these tools by all researchers, emphasizing the indispensable role of human agency in
research outcomes. It is imperative for researchers to utilize these tools with a nuanced
understanding that the responsibility for the outcome lies with the researcher rather than
the machine.
Articles presenting technological tools for solving problems are characteristic of
modern technologies and appear in the literature 7-12. They are particularly important at an
early stage of the development of a technology that is potentially useful in a specific
field11. Such articles emphasize the significant impact that AI tools can have on
optimizing the systematic review process. 13 By leveraging AI tools, researchers can
enhance efficiency in academic writing, refine search strategies, screen relevant
literature, extract data, and summarize findings, thereby streamlining the entire
systematic review process13. This optimization not only saves time but also ensures a
more thorough and comprehensive review of existing literature.

2. Theoretical background

2.1. Science automation and artificial intelligence

Science automation represents a pivotal paradigm shift in scientific research,
incorporating advanced technologies, such as robotics, AI, machine learning, and other
advanced technologies, to streamline and optimize various scientific processes 14. This
paradigm spans multiple domains, including computer science, mathematics, health
service research, data science, engineering, and management science. In particular, the
fusion of AI and automation has given rise to novel tools and methodologies that can
significantly enhance the efficiency, accuracy, and accessibility of scientific research. 14
The rationale for automating systematic literature reviews (SLR) is their strongly
increasing number and extraordinary workload.
AI is the development of computer systems that can perform tasks that typically
require human intelligence. These tasks include learning, reasoning, problem solving,
perception, and language understanding15-16. AI systems are designed to process and
analyze large amounts of data, identify patterns, and make predictions or decisions based
on that data17. They have the potential to transform various aspects of human life,
including healthcare, education, logistics, and business efficiency. 18-21
AI has become increasingly relevant in academic research, with its applications and
implications explored in various fields. AI has the potential to accelerate the research
process and support researchers in multiple tasks. It is perceived as helpful in information
gathering, other narrow tasks, and supporting impact and interdisciplinarity. 22 However,
4 Tomczyk, Brüggemann, & Doligalski

there is a concern that using AI to “accelerate” research processes may contribute to the
negative aspects of academic culture. They should be seen as tools to assist, rather than
replace, human creativity.22

2.2. Automation of systematic literature reviews

Systematic literature review is a research method that enables the identification,
selection, critical evaluation, and synthesis of existing literature in a rigorous, transparent,
and repeatable manner, leading to robust conclusions about what is known and what is
not known in peer-reviewed research articles. 23 It is used in scientific journals in various
fields.24 According to Scopus, in 2023, 2773 scientific articles in management,
accounting, and finance were published, in which this method was mentioned. Compared
to 2018 in 2022, 1638 in 2021, and 1408 in 2020, an upward trend close to exponential
can be seen.
Automating systematic literature reviews in the management sciences offers several
benefits. First, it allows a more efficient and timely review process. Conducting a manual
systematic literature review can take anywhere from six months to several years for a
full-time researcher.12 Automation tools can help speed up the process while maintaining
the high standards of a systematic review.12
Second, automation can improve the accuracy and reliability of review processes.
Manual reviews are prone to human error, such as overlooking relevant studies or
misinterpreting the data. Automation tools can help minimize these errors by using
machine learning techniques to identify and extract relevant information from a large
literature volume.25 This can lead to more comprehensive and reliable reviews.
Third, automation can enhance the reproducibility and transparency of systematic
reviews. Using automated tools, researchers can document and track each step of the
review process, making it easier for others to replicate the study and verify the findings. 26
This promotes transparency and allowing for more rigorous evaluation of the review
methodology.
Apart from the benefits of automating systematic literature reviews, there are also
several doubts and problems. We discuss them in the fourth section of this article.
Based on the procedure of Vrontis and Christofi 24, we identified the following stages
in a systematic literature review:

(i) formulating a review question23-24,27-28,

(ii) data collection with keyword identification23,28,
(iii) database preparation24,27,
(iv) bibliometric analysis and descriptive statistics24,28-29,
(v) data extraction and synthesis24,28-29,
(vi) report preparation23,24.

The role of AI in formulating a research question in a systematic literature review is

multi-faceted. AI can assist in the initial stages of a systematic literature review by
automating the process of identifying research questions based on an analysis of existing
5 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

literature. AI-enabled tools, such as ChatGPT, can methodically and productively support
the identification of research questions, enhancing the efficiency and effectiveness of the
scoping process.30 By leveraging machine learning algorithms, AI can also analyze a vast
amount of literature to identify gaps, trends, and emerging topics, thereby guiding the
formulation of relevant research questions. 31-33 However, it is essential to note that while
AI can significantly contribute to identifying research questions, ethical considerations,
and biases in machine learning models must be carefully addressed to ensure integrity
and fairness.34
In the data collection and keyword identification stage, AI technologies, including
machine learning and natural language processing, have the potential to automate and
streamline the data collection process, thereby enhancing the efficiency and accuracy of
systematic literature reviews.35-36 AI can automate data extraction, making the systematic
review process more thorough and comprehensive. 37 In addition, AI-driven technologies
can contribute to data collection, analysis, and modelling in a variety of domains. 38
However, it is important to consider the potential challenges and ethical implications of
AI-driven data collection in systematic literature reviews. Risks such as data
manipulation, privacy concerns, and limitations in data access and transparency need to
be addressed to ensure AI’s responsible and ethical use of AI in data collection. 39
In modern data management, AI plays a significant role in automating data
extraction, cleaning, and organization, enhancing the efficiency and accuracy of database
preparation processes.40 By leveraging AI, researchers can conduct systematic literature
reviews using a data-driven approach and quantitative methodology, providing an
overview of the emerging intellectual structure of AI research. 41 Furthermore, AI
facilitates the identification of research trends in the field through systematic bibliometric
literature reviews.42
The role of AI in the bibliometric preparation for a systematic literature review is not
as significant as in the previous stages. AI has been utilized in bibliometric analysis to
comprehensively analyze the existing literature, identify trends, and provide insights.
Bibliometric analysis involves using visual software, such as CiteSpace or SystemPro, to
use particular AI implementations. As a rule, bibliometric analysis is supported by
dedicated software that is rarely based on AI, such as VOSviewer or Sci-Mat.
The role of AI in data extraction and synthesis in a systematic literature review is
significant and has been increasingly recognized by researchers. AI has been utilized to
automate the literature screening process, extract data, and analyze research themes 43 and
to process large volumes of data, enabling the evaluation of extensive datasets. 44-45 In the
context of thematic analysis, AI has been employed to identify and analyze themes in
qualitative research, contributing to the systematic analysis of literature and the
generation of analytical themes.46-47
The concluding stage of the systematic literature review process–report preparation–
presents another pivotal arena for the application of AI. In this final phase, AI’s
capabilities extend to meticulous screening and synthesis of the literature, ensuring that
the most relevant and impactful studies are highlighted in the report. AI algorithms have
6 Tomczyk, Brüggemann, & Doligalski

been applied to automate the literature screening procedure in systematic reviews,

demonstrating the potential of AI to streamline the initial stages of report preparation.
Furthermore, AI plays an important role in identifying and elucidating emerging
intellectual structures in the field of study.

3. Evaluation of AI-based tools for systematic literature reviews

To answer our research question, we identified the AI applications available in the
market. Those that appeared on theresanaiforthat.com after a search for the keywords
‘scientific research’ and ‘academic research’ were considered as such. We identified 15
such applications. We have supplemented this set with the other seven applications that
we were able to identify in further research. Finally, we identified 22 applications.
We assessed the applications in terms of the stages described above. For each stage,
we evaluated each application with respect to subsequent stages of systematic literature
review.
We created a list of applications on April 17, 2024. Tab. One presents a matrix of all
the applications analyzed and their assigned functions. The rows contain the names of the
applications and columns of the stages of the systematic literature review (i.e., stages (i)
to (vi)). Furthermore, we report the “PDF uploading ability”. In this column, “1” means
that the application can analyze one document at a time and “1+” means that it can
analyze more of them simultaneously.
Overall, the applications described were diverse in terms of the characteristics
presented, as shown in Tab. 1. This means that the market for applications that support
researchers performing literature reviews is diverse.
The applications analyzed can be summarized based on their ranking in the “SUM”
column: Petal is the most comprehensive research application offering possibilities for
data analysis, database access, extraction, reporting, and uploading more than one PDF
file. In contrast, Textero users are provided with the same possibilities, except that only
one PDF file can be uploaded. MirrorThink and Consensus.app offer broad possibilities
but have no PDF file opportunities. All further AI applications analyzed provide
comparatively limited capabilities for AI-based systematic literature reviews across the
stages under consideration.
Notably, none of the applications listed encompass bibliometric analysis
functionalities. Such capabilities are predominantly embedded within science mapping
software.48-49 Additional analytical prospects within the scope of a systematic literature
review can be achieved through the versatile approach of variable science mapping. 50
Subsequently, the ensuing section of this article discusses the utilization of AI
applications across various stages within systematic literature reviews.
International Journal on Artificial Intelligence Tools
Vol. XX, No. X (2025) 1–17
 World Scientific Publishing Company

Table 1. The relevance of AI applications at different stages of the systematic literature review

Stages of systematic literature review PDF

b) data collection d) bibliometric e) data SU uploa
a) formulating a c) database f) report
with keyword analysis and extraction and M d
review question preparation preparation
identification descriptive statistics synthesis ability
Petal X X X X X 5 1+
Epsilon X X X X X 5 1+
Textero X X X X X 5 1
MirrorThink X X X X X 5
Scite X X X X X 5 1+
Consensus.app X X X X X 5
Research Buddy X X X X 4
Perplexity X X X X 4
Jenni X X X X 4 1+
Elicit X X X 3 1+
SciSpace X X X 3 1+
Open Read X X X 3 1
Semantic Scholar X X X 3
SystemPro X X X 3
Cambiran X X X 3
Paper Digest X X X 3
PowerDrill X X 2 1+
AcademicGPT X X 2 1
Julius X X 2
SciSummary X 1 1
Scholarcy X 1 1
ResearchAIde X 1 1
SUM 6 16 16 0 22 12

1
International Journal on Artificial Intelligence Tools
Vol. XX, No. X (2025) 1–17
 World Scientific Publishing Company

The application analysis can be summarized based on their ranking in the “SUM”
column as follows: Petal, Epsilon, and Textero are the top-ranked apps, offering
comprehensive research features encompassing data analysis, database access, extraction,
reporting, and PDF management. MirrorThink, SciSpace, Open Read, Scite, Elicit, and
Consensus.app closely follow, excelling in various research aspects including database
access and extraction. Research Buddy, Academic GPT, Perplexity, Power Drill, and
Jenni have moderate rankings, focusing on specific research needs such as database
access and content generation. Semantic Scholar, SystemPro, SciSummary, Scholarcy,
and ResearchAIde are lower-ranked, emphasizing, and report generation. Cambiran and
Paper Digest provided similar data and access to databases. Julius has the lowest rank,
specializing in, with limited coverage in other research areas.
The applications described are diverse in terms of the characteristics presented, as
evidenced by the coefficient of variation read from the last row: Vc = 60%. This means
that the market for applications that support researchers performing literature reviews is
diverse.
It is worth noting that there is no single application in the list that performs a
bibliometric analysis. Such functions are primarily contained in science mapping
software.42-43 In the following part of the article, the article explains the benefits and
determinants of the use of AI at different stages of the systematic literature review.
However, the presented quantitative analysis does not reflect the qualitative values of
the presented applications. Therefore, it is worth noting that the key characteristics of the
presented solutions are divided into groups according to the phases of the systematic
literature review.
Support at the stage of formulating a research question consists of identifying the
existing relationships between variables. The best way to do this is with the consensus
app, which provides a fraction of positive and negative responses resulting from the
analyzed articles to the question asked, for example, whether there is a relationship
between customer satisfaction and customer loyalty. Other applications, either directly
(Scite) or indirectly (Epsilon), answer this question.
Support in the field of data retrieval consists of identifying articles relevant to a given
area. These articles are most often searched in the Semantic Scholar database and sorted
by importance. This search is similar to searching traditional databases, such as Scopus
and Web of Science. The difference, however, is that in this case, you do not need to
enter keywords but only formulate a question.
Building a database was based on the presentation of searched articles in the form of a
list. However, some applications allow you to export the list to a file format for your
choice (for example,. csv, BIB text or RIS). For example, elicit and cispace have this
feature.
Data extraction was possible for each application presented in the table. Extraction
refers to the ability to extract key information from metadata according to a specific
algorithm. However, it should be noted that extraction is particularly effective for

1
2 Tomczyk, Brüggemann, & Doligalski

applications that allow data to be exported to a database. In this situation, the Elicit and
Scispace applications are again the best to use, as they allow you to export to the CSV
format. A file that has been appropriately processed in Excel is significant for this
purpose.
In the writing of a report, applications that write ready-made text based on the
question asked and then paraphrase it are particularly useful. Applications of ChatDOC
and Epsilon are extremely useful in this regard.

4. Artificial intelligence technical usage limitation

This chapter presents the key limitations of the use of AI in scientific writing along
with their polemics.

The last two steps in preparing an SLR are data extraction and report-writing. These
stages are particularly sensitive to data quality. 51 For this reason, we recommend the use
of applications based on articles selected during the data collection and database
preparation stages. Thanks to this, the researcher can control the generated content and
check it on an ongoing basis. This is how ChatDOC application works, for example.
Another issue is the understanding of highly specialized and nuanced languages. This
is part of the stage of writing the report, during which it is particularly important for
scientists to control the results of the algorithm’s work on an ongoing basis. It should be
remembered that the scientist controls a process that cannot be completely automatic at
this point.52 Although modern algorithms are getting better at dealing with specialized
language, human control still seems to be indispensable at this stage.
Another problem is context understanding, ambiguity resolution, and figurative
language handling. This is caused by limitations in current NLP technologies, and can
lead to a decrease in AI’s effectiveness of AI. It should be noted that modern technologies
are improving in these areas. 53 Understanding the context still remains the main domain
of the scientist, and in this respect, its assessment plays a decisive role. 52
Some authors question the credibility of articles created by AI, arguing that they act
as a blackbox.54 However, in the case of writing a scientific article using ChatDOC or
Epsilon, the scientist has the opportunity to verify and compare the original text with the
text prepared by the machine at every stage. Therefore, even if we are dealing with a
black box, the results of this study are relatively easy to verify.
Finally, the problem of multidisciplinary literature arises. 55 Due of the above-
mentioned problem of contextual analysis, the result of the algorithm’s work must be
subject to strict control by the researcher.

5. Discussion of implications for research

As our research shows, AI applications are useful for conducting systematic literature
reviews. This type of review involves seven stages, six of which are AI support
researchers. There are many implications for researchers and the science itself.
The term used in this study is science automation. This term and similar terms have
appeared for several years in various contexts. 56 DARPA attempts to create an
3 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

autonomous scientist, an AI agent, which (not who) would be capable of sceptical

learning and reasoning to generate creative hypotheses and experiments. 57
According to the Cambridge Dictionary, automation is the use of machines and
computers that can operate without the need for human control. 58 We argue that using AI
applications in a systematic literature review process requires a high level of researcher
involvement, including the setting of boundary conditions, and the subsequent critical
analysis of the results of the application’s actions and their interpretation. AI applications
in a systematic literature review have a different function from calculators or
spreadsheets in the case of calculations. This is because producing a systematic literature
review has many of the characteristics of an unstructured problem that does not have a
single correct solution, but rather multiple sets of solutions and outcomes. 59 Conducting a
systematic literature review requires evaluation, nuance, critical judgement, and decision-
making, often based on a deep, long-term understanding of a particular field of
knowledge. The current developments in AI do not ensure that these tasks are performed.
Therefore, the key tasks in the research process are carried out by the researcher and not
the algorithm.
In the context of generative AI, there is often talk about augmenting the skills of
employees performing cognitive tasks.60 In other is, AI tools can enhance or extend
human capabilities in various tasks or processes. In our view, this approach also fits the
use of AI in a systematic literature review. AI applications act as assistants that perform
defined activities, which are the essence of a given step in the process of producing a
systematic literature review. Decisions that require critical reflection, value judgements,
or are closely related to the desired outcome are made by the researchers themselves. By
using AI applications, researchers can save time and effort that can be spent on most
value-added tasks. The use of AI applications can lead to higher-quality systematic
literature reviews if the technology is correctly used.
It is worth mentioning a certain danger associated with the facilitation of systematic
literature review. This may lead to an increase in the popularity of this type of research,
especially because it does not require external funding. This may accelerate the
unfavorable phenomenon of publishing many publications of questionable quality. This
phenomenon is evidenced by data according to which, of the articles published in 2017-
2019 in Polish scientific journals, at least one citation in 2020 was obtained by only
approximately 7% of articles.61
Returning to the keyword in the title, the automation of science, it should be noted
that systematic literature reviews are one of many types of scientific articles. A type of
scientific article similar to a systematic literature review was a meta-analysis. Both types
combine and analyze data from studies on similar research topics. 62 However, there are
differences between them. A meta-analysis was performed to combine the statistical
results of the different studies. A systematic literature review presented, collected, and
structured the results of several studies. Therefore, the meta-analysis used statistical
analyses of quantitative data from different studies. The AI applications used for
systematic literature reviews rely on large language models, that is, software based on the
4 Tomczyk, Brüggemann, & Doligalski

analysis of language rather than numbers. Different classes of software help researchers
conduct meta-analyses and systematic literature reviews. Evaluating how these classes of
software contribute to automating and augmenting researchers for different types of
scientific problems could be the subject of further study.
In addition, there are articles that are empirical research reports or strictly conceptual. In
the former case, it is necessary to conduct one’s own research and not rely on the content
of other articles. AI applications can play a supporting role to an extent that depends very
much on the specifics of the study in question. At the other end of the continuum are
conceptual papers, which, according to Gilson and Goldberg 63, develop logical and
complete arguments for associations rather than empirically testing them. As a result,
conceptual papers do not have data; they integrate and propose relationships among
constructs. In other words, they developed and proposed new theories for a given field.
Advances in AI do not guarantee automation of theoretical development.
An issue related to science automation is the use of AI in the preparation of students’
written work, especially dissertations. Some universities have also introduced regulations
in this area. One example is the SGH Warsaw School of Economics, which specified the
use of AI in ten areas 64. These are ideation, knowledge acquisition and literature review,
writing, text operations, graphics creation, programming, data analysis, economic and
mathematical modelling, AI as a research subject, and reporting on the use of AI. The
variety of highlighted areas reflects the diversity of ways in which AI can be applied to
cognitive work. As stated in the regulations, full responsibility for the content of the
written work rests with the author. AI can serve as a tool; however, it does not replace
critical thinking, self-reflection, or a creative approach to the subject. Therefore, AI
serves to augment cognitive work. The automation of certain tasks was also allowed.
However, these are not value-adding activities but those that must be performed, such as
formatting footnotes and bibliographies, correcting style, or clarity of the text. This
approach was consistent with the results of the present study. AI applications can serve as
assistants, but humans perform the most important value-adding activities.

6. Conclusion, limitations and further research directions

In conclusion, researchers conducting systematic literature reviews in the management
sciences have many AI tools. The choice of an application should be based on the
specific requirements of the research project, as these tools offer varying degrees of
support across different stages of the systematic literature review process. In addition, for
bibliometric analysis, researchers may need to explore specialized software beyond the
applications listed in this study.
The main limitation of this study is the subjective classification of the software
features presented. In the case of more detailed stages of a systematic literature review,
an analysis by competent judges is necessary. An additional limitation is the lack of
assessment of functionality. It is obvious that applications that perform data extraction do
so at different levels of sophistication. Moreover, the analysis does not include
applications for ancillary activities, such as copywriting or grammar checking.
5 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

We recommend that researchers conducting systematic literature reviews in the

management sciences perform qualitative assessments of tested AI applications to
thoroughly differentiate their functionalities. Tomczyk et al. provided an initial
comparison between AI-based applications and conventional literature searches. 65
Exploring innovative AI-powered approaches that separate and automate various SLR
stages can significantly enhance research efficiency. Leveraging AI tools such as Textero
or Epsilon for idea generation and source searching can streamline the research process.
Furthermore, researchers should consider using AI to enhance narrative writing and
optimize article structure. Staying informed about evolving AI technologies and being
open to adapting research methodologies will keep researchers at the forefront of
scientific innovation and ultimately improve research quality and productivity. It is worth
noting that the use of AI can lead to a dominant role for software over humans,
previously resulting in a kind of “artificial subjectivity” of the review and a lack of
control of the author over its results. In this situation, developing a “separated AI
systematic literature review pathway” strategy for working with AI in this area is
necessary.
This research demonstrates that AI is still in its early stages, but holds great potential
for scientific writing (see Table 1 and related discussion). However, serious threats still
exist. Only when researchers learn to properly utilize these new possibilities and
understand their limitations can AI revolutionize the future of academic science.
Ultimately, the researcher was responsible for the outcome, not the AI. Therefore, every
user of AI tools should not rely on them blindly, but use them critically and reflectively.
Nevertheless, AI tools have the potential to revolutionize the academic world, but only if
used correctly.

References
1. C. Loebbecke and A. Picot (2015). Reflections on societal and business model transformation
arising from digitization and big data analytics: A research agenda. The Journal of Strategic
Information Systems, 24(3), 149-157. https://doi.org/10.1016/j.jsis.2015.08.002
2. J. Siderska (2020). Robotic Process Automation—a driver of digital transformation?.
Engineering Management in Production and Services, 12(2), 21-31.
https://doi.org/10.2478/emj-2020-0009
3. G. Chryssolouris, D. Mavrikios, N. Papakostas, D. Mourtzis, G. Michalos and K. Georgoulias
(2009). Digital manufacturing: history, perspectives, and outlook. Proceedings of the
Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, 223(5),
451-462. https://doi.org/10.1243/09544054JEM1241
4. T. De Bie, L. De Raedt, J. Hernández-Orallo, H. H. Hoos, P. Smyth and C. K. Williams
(2022). Automating data science. Communications of the ACM, 65(3), 76-87.
https://doi.org/10.48550/arXiv.2105.05699
5. T. Davenport, A. Guha, D. Grewal and T. Bressgott (2020). How artificial intelligence will
change the future of marketing. Journal of the Academy of Marketing Science, 48, 24-42.
https://doi.org/10.1007/s11747-019-00696-0
6 Tomczyk, Brüggemann, & Doligalski

6. Z. Xi, W. Chen, X. Guo, W. He, Y. Ding, B. Hong, ... and T. Gui (2023). The rise and
potential of large language model based agents: A survey.
https://doi.org/10.48550/arXiv.2309.07864
7. L. Li, N. Lv and W. Li (2022). Research on application of graph neural network in water
quality prediction. International Journal on Artificial Intelligence Tools, 31(01), 2250018.
8. B. Bekiroglu and B. Korel (2023). Neural Network-based Tool for Survivability Assessment
of K-variant Systems. International Journal on Artificial Intelligence Tools, 32(04), 2350049.
9. K. Davies, C.M. Keet and A. Lawrynowicz (2019). More effective ontology authoring with
test-driven development and the TDDonto2 tool. International Journal on Artificial
Intelligence Tools, 28(07), 1950023.
10. H. Khalil, D. Ameen and A. Zarnegar (2022). Tools to support the automation of systematic
reviews: a scoping review. Journal of Clinical Epidemiology, 144, 22-42.
11. A. Correia, A. Grover, S. Jameel, D. Schneider, P. Antunes and B. Fonseca (2023). A hybrid
human–AI tool for scientometric analysis. Artificial Intelligence Review, 56(Suppl 1), 983-
1010.
12. A. J. Van Altena, R. Spijker and S. D. Olabarriaga (2019). Usage of automation tools in
systematic reviews. Research synthesis methods, 10(1), 72-82.
https://doi.org/10.1002/jrsm.1335
13. H. Fleischer, K. Ramani, K. Blitti, T. Roddelkopf, M. Warkentin, D. Behrend and K. Thurow
(2018). Flexible automation system for determination of elemental composition of
incrustations in clogged biliary endoprostheses using ICP-MS. SLAS TECHNOLOGY:
Translating Life Sciences Innovation, 23(1), 83-96.
https://doi.org/10.1177/2472630317727451
14. G. Bansal, T. Wu, J. Zhou, R. Fok, B. Nushi, E. Kamar, ... and D. Weld (2021). Does the
whole exceed its parts? the effect of ai explanations on complementary team performance.
In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-
16). https://doi.org/10.1145/3411764.3445717
15. L. David and K. Alan (2010). Artificial intelligence: foundations of computational agents.
Choice Reviews Online, 48(04), 48-2130-48-2130. https://doi.org/10.5860/choice.48-2130
16. M. Janssen, P. Brous, E. Estevez, L. S. Barbosa and T. Janowski (2020). Data governance:
Organizing data for trustworthy Artificial Intelligence. Government Information
Quarterly, 37(3), 101493. https://doi.org/10.1016/j.giq.2020.101493
17. S. Al’Aref, K. Anchouche, G. Singh, P. Slomka, K. Kolli, A. Kumar, … and J. Min (2018).
Clinical applications of machine learning in cardiovascular disease and its relevance to
cardiac imaging. European Heart Journal, 40(24), 1975-1986.
https://doi.org/10.1093/eurheartj/ehy404
18. P. Radanliev, D. Roure, C. Maple and O. Santos (2022). Forecasts on future evolution of
artificial intelligence and intelligent systems. Ieee Access, 10, 45280-45288.
https://doi.org/10.1109/access.2022.3169580
19. X. Zhu, Y. Shi and N. Liu (2022). Artificial intelligence technology in modern logistics
system. International Journal of Technology Policy and Management, 22(1/2), 66.
https://doi.org/10.1504/ijtpm.2022.10046969
20. M. Artar and P. ERDİL (2022). Artificial intelligence in project management: an application
in the banking sector. Akademik Araştırmalar Ve Çalışmalar Dergisi (Akad), 14(27), 323-
334. https://doi.org/10.20990/kilisiibfakademik.1159862
21. J. Chubb, P. Cowling and D. Reed (2022). Speeding up to keep up: exploring the use of AI in
the research process. AI & society, 37(4), 1439-1457. https://doi.org/10.1007/s00146-021-
01259-0
7 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

22. M. Christofi, D. Vrontis and J. W. Cadogan (2021). Micro-foundational ambidexterity and

multinational enterprises: a systematic review and a conceptual framework. International
Business Review, 30(1), 101625. https://doi.org/10.1016/j.ibusrev.2019.101625
23. D. Vrontis and M. Christofi (2021). R&D internationalization and innovation: A systematic
review, integrative framework and future research directions. Journal of Business
Research, 128, 812-823. https://doi.org/10.1016/j.jbusres.2019.03.031
24. R. van Dinter, B. Tekinerdogan and C. Catal (2021). Automation of systematic literature
reviews: A systematic literature review. Information and Software Technology, 136, 106589.
https://doi.org/10.1016/j.infsof.2021.106589
25. G. Tsafnat, P. Glasziou, M. K. Choong, A. Dunn, F. Galgani and E. Coiera (2014). Systematic
review automation technologies. Systematic reviews, 3, 1-15.
http://www.systematicreviewsjournal.com/content/3/1/74
26. E. Leonidou, M. Christofi, D. Vrontis and A. Thrassou (2020). An integrative framework of
stakeholder engagement for innovation management and entrepreneurship
development. Journal of Business Research, 119, 245-258.
https://doi.org/10.1016/j.jbusres.2018.11.054
27. K. E. Mcquade, C. Harrison and H. Tarbert (2021). Systematically reviewing servant
leadership. European Business Review, 33(3), 465-490. https://doi.org/10.1108/EBR-08-2019-
0162
28. D. Siemieniako, M. Mitręga and K. Kubacki (2022). The antecedents to social impact in inter-
organizational relationships–A systematic review and future research agenda. Industrial
Marketing Management, 101, 191-207. https://doi.org/10.1016/j.indmarman.2021.12.014
29. B. Burger, D. Kanbach, S. Kraus, M. Breier and V. Corvello (2023). On the use of ai-based
tools like ChatGPT to support management research. European Journal of Innovation
Management, 26(7), 233-241. https://doi.org/10.1108/ejim-02-2023-0156
30. J. Schaefer, M. Lehne, J. Schepers, F. Prasser and S. Thun (2020). The use of machine
learning in rare diseases: a scoping review. Orphanet Journal of Rare Diseases, 15(1).
https://doi.org/10.1186/s13023-020-01424-6
31. A. Lejeune, A. Glaz, P. Perron, J. Sebti, E. Baca-García, M. Walter … and S. Berrouiguet
(2022). Artificial intelligence and suicide prevention: a systematic review. European
Psychiatry, 65(1). https://doi.org/10.1192/j.eurpsy.2022.8
32. F. Kharbat, A. AlShawabkeh and M, Woolsey (2020). Identifying gaps in using artificial
intelligence to support students with intellectual disabilities from education and health
perspectives. Aslib Journal of Information Management, 73(1), 101-128.
https://doi.org/10.1108/ajim-02-2020-0054
33. Y. Zhang and R. Zhu (2022). A brief review on algorithmic fairness. Management System
Engineering, 1(1). https://doi.org/10.1007/s44176-022-00006-z
34. N. Murray, M. Unberath, G. Hager and F. Hui (2019). Artificial intelligence to diagnose
ischemic stroke and identify large vessel occlusions: a systematic review. Journal of
Neurointerventional Surgery, 12(2), 156-164. https://doi.org/10.1136/neurintsurg-2019-
015135
35. B. Halamoda-Kenzaoui, R. Etienne, J. Piovesan, A. Puertas-Gallardo and S. Bremer‐
Hoffmann (2021). Toxic effects of nanomaterials for health applications: how automation can
support a systematic review of the literature?. Journal of Applied Toxicology, 42(1), 41-51.
https://doi.org/10.1002/jat.4204
36. S. Jonnalagadda, P. Goyal and M. Huffman (2015). Automating data extraction in systematic
reviews: a systematic review. Systematic Reviews, 4(1). https://doi.org/10.1186/s13643-015-
0066-7
8 Tomczyk, Brüggemann, & Doligalski

37. Z. Doborjeh, N. Hemmington, M. Doborjeh and N. Kasabov (2021). Artificial intelligence: a

systematic review of methods and applications in hospitality and tourism. International
Journal of Contemporary Hospitality Management, 34(3), 1154-1176.
https://doi.org/10.1108/ijchm-06-2021-0767
38. C. Trocin, P. Mikalef, Z. Papamitsiou and K. Conboy (2021). Responsible ai for digital health:
a synthesis and a research agenda. Information Systems Frontiers.
https://doi.org/10.1007/s10796-021-10146-4
39. P. Dhamija and S. Bag (2020). Role of artificial intelligence in operations environment: a
review and bibliometric analysis. The TQM Journal, 32(4), 869-896.
https://doi.org/10.1108/tqm-10-2019-0243
40. M. Mariani, R. Perez‐Vega and J. Wirtz (2021). Ai in marketing, consumer research and
psychology: a systematic literature review and research agenda. Psychology and Marketing,
39(4), 755-776. https://doi.org/10.1002/mar.21619
41. R. Raimundo and A. Rosário (2021). The impact of artificial intelligence on data system
security: a literature review. Sensors, 21(21), 7029. https://doi.org/10.3390/s21217029
42. Y. Zhang, S. Liang, Y. Feng, Q. Wang, F. Sun, S. Chen … and P. Hui. (2020). A research
protocol for a systematic review of automatic literature screening in medical evidence
synthesis. https://doi.org/10.21203/rs.3.rs-62316/v1
43. A. Grothen, B. Tennant, C. Wang, A. Torres, B. Sheppard, G. Abastillas … and D. Rivera
(2020). Application of artificial intelligence methods to pharmacy data for cancer surveillance
and epidemiology research: a systematic review. Jco Clinical Cancer Informatics, (4), 1051-
1058. https://doi.org/10.1200/cci.20.00101
44. Y. Zhang, S. Liang, Y. Feng, Q. Wang, F. Sun, S. Chen … and H. Pan (2022). Automation of
literature screening using machine learning in medical evidence synthesis: a diagnostic test
accuracy systematic review protocol. Systematic Reviews, 11(1).
https://doi.org/10.1186/s13643-021-01881-5
45. M. Lally (2022). Understanding the experiences of fulbright visiting scholars—a qualitative
systematic review. Education Sciences, 12(2), 90. https://doi.org/10.3390/educsci12020090
46. C. Maley, N. Pagana, C. Velenger and T. Humbert (2016). Dealing with major life events and
transitions: a systematic literature review on and occupational analysis of spirituality.
American Journal of Occupational Therapy, 70(4), 7004260010p1-7004260010p6.
https://doi.org/10.5014/ajot.2016.015537
47. C. Chen (2017). Science mapping: a systematic review of the literature. Journal of data and
information science, 2(2), 1-40. https://doi.org/10.1515/jdis-2017-0006
48. J. A. Moral-Muñoz, E. Herrera-Viedma, A. Santisteban-Espejo and M. J. Cobo (2020).
Software tools for conducting bibliometric analysis in science: An up-to-date
review. Profesional de la Información, 29(1). https://doi.org/10.3145/epi.2020.ene.03
49. P. Tomczyk, P. Brüggemann and J. Paul (2024). Variable science mapping as literature review
method. Journal of Marketing Analytics, 1-13. https://doi.org/10.1057/s41270-024-00336-9
50. C. Norman (2020). Systematic review automation methods. Information Retrieval [cs.IR].
Université Paris-Saclay; Universiteit van Amsterdam, 2020. English. fNNT : 2020UPASS028
51. M. Khalifa, and M. Albadawy (2024). Using artificial intelligence in academic writing and
research: An essential productivity tool. Computer Methods and Programs in Biomedicine
Update, 100145. https://doi.org/10.1016/j.cmpbup.2024.100145
52. H.O. Khogali and S. Mekid (2023). The blended future of automation and AI: Examining
some long-term societal and ethical impact features. Technology in Society, 73, 102232.
https://doi.org/10.1016/j.techsoc.2023.102232
53. L. Messeri and M.J. Crockett (2024). Artificial intelligence and illusions of understanding in
scientific research. Nature, 627(8002), 49-58. https://doi.org/10.1038/s41586-024-07146-0
9 Instructions for Typing Manuscripts (The Automation of Science? Possibilities and Boundaries of AI
Applications for Conducting Systematic Literature Reviews)

54. M. Sartas, S. Cummings, A. Garbero and A. Akramkhanov (2021). A human machine hybrid
approach for systematic reviews and maps in international development and social impact
sectors. Forests, 12(8), 1027. https://doi.org/10.3390/f12081027
55. R. D. King, J. Rowland, S. G. Oliver, M. Young, W. Aubrey, E. Byrne ... and A. Clare (2009).
The automation of science. Science, 324(5923), 85-89.
https://doi.org/10.1126/science.1165620
56. DARPA (2023). Inside DARPA’s search for an 'autonomous scientist' to support its
researchers, https://www.nextgov.com/artificial-intelligence/2023/11/inside-darpas-search-
autonomous-scientist-support-its-researchers/392139/
57. Cambridge University Press. (n.d.). Entry “Automation” In Cambridge Dictionary. Retrieved
from https://dictionary.cambridge.org/dictionary/english/automation
58. M. Rosenberg (2011), IESE’s Six Step Process for Resolving Unstructured Problems,
https://blog.iese.edu/rosenberg/files/2011/12/Six-Step-Process-for-Resolving-Unstructured-
Problems.pdf
59. MIT (2024). MIT launches Working Group on Generative AI and the Work of the Future,
https://news.mit.edu/2024/mit-launches-working-group-generative-ai-and-work-of-the-future-
0328
60. Biała Księga. Raport na temat rynku wydawnictw naukowych w Polsce w 2023 roku (2023),
https://forumakademickie.pl/wp-content/uploads/2023/12/Biala-ksiega_2023.pdf
61. E. Ahn and H. Kang (2018). Introduction to systematic review and meta-analysis. Korean
journal of anesthesiology, 71(2), 103-112. https://doi.org/10.4097/kjae.2018.71.2.103
62. L. L. Gilson and C. B. Goldberg (2015). Editors’ comment: so, what is a conceptual paper?.
Group & Organization Management, 40(2), 127-130.
https://doi.org/10.1177/1059601115576425
63. Principles of using artificial intelligence in the preparation of written assignments at the SGH
Warsaw School of Economics. (2024). SGH Warsaw School of Economics.
https://www.researchgate.net/publication/378487927_Principles_of_using_artificial_intelligen
ce_in_the_preparation_of_written_assignments_at_the_SGH_Warsaw_School_of_Economics
/references
64. P. Tomczyk, P. Brüggemann, N. Mergner, M. Petrscu (2023). Exploring AI’s Role in
Literature Searching: Traditional Methods Versus AI-Based Tools in Analyzing Topical E-
Commerce Themes, in: F. J. Martinez-Lopez, L. F. Martinez, and P. Brüggemann (Ed.),
Advances in Digital Marketing and eCommerce – Fifth International Conference (Springer,
Cham, 2024), pp. 141-148.