See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/384359784

Unearthing AI coaching chatbots capabilities for professional coaching: a

systematic literature review

Article in Journal of Management Development · September 2024

DOI: 10.1108/JMD-06-2024-0182

CITATION READS

1 545

2 authors:

Lidia Plotkina Subramaniam Sri Ramalu

Northern University of Malaysia Northern University of Malaysia
2 PUBLICATIONS 5 CITATIONS 77 PUBLICATIONS 1,199 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Subramaniam Sri Ramalu on 30 September 2024.

The user has requested enhancement of the downloaded file.

 The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0262-1711.htm

Journal of
Unearthing AI coaching chatbots Management
Development
capabilities for professional
coaching: a systematic
literature review
Lidia Plotkina and Subramaniam Sri Ramalu Received 5 June 2024
Revised 20 August 2024
College of Business, Universiti Utara Malaysia, Sintok, Malaysia Accepted 5 September 2024

Abstract
Purpose – Recent advances in coaching technology enhanced its accessibility and affordability for a broader
population. In the imposing growth of economy and the demand for extensive coaching intervention for
executives, artificial intelligence (AI)-based coaching is one of the possible solutions. While the evidence of AI
coaching effectiveness is expanding, a comprehensive understanding of the field remains elusive. In
particular, the true potential of AI coaching tools, ethical considerations and their current functionality are
subjects of ongoing investigation.
Design/methodology/approach – The systematic literature review was conducted to extract experimental
results and concepts about utilizing AI in coaching practice. The paper presents the primary capabilities of
state-of-the-art coaching tools and compares them with human coaching.
Findings – The review shows that AI coaching chatbots and tools are effective for narrow tasks such as goal
attainment, support for various psychological conditions and induction of reflection processes. Whereas, deep
long-term coaching, working alliance and individualized approach are out of current AI coaching competence.
In the current state, AI coaching tools serve as complementary helping tools that cannot replace human
coaching. However, they have the potential to enhance the coach’s performance and serve as valuable
assistants in intricate coaching interventions.
Originality/value – The review offered insights into the current capabilities of AI coaching chatbots, aligned
with International Coaching Federation set of competencies. The review outlined the drawbacks and benefits
of chatbots and their areas of application in coaching.
Keywords Executive coaching, Coaching chatbot, Virtual coach, AI coaching, Coaching effectiveness
Paper type Literature review

1. Introduction
Coaching practice is transforming drastically and moving into more digital forms of
interaction. Taking into account the rising need for cost-effective and affordable coaching in
the growing market reality, artificial intelligence (AI) coaching and distance coaching, which
uses various technologies, also recognized as e-coaching or virtual coaching could be a
possible solution. The massive growth of technological interventions in executive coaching
appeared due to the plethora of factors and include the simplification of logistics, cost-
savings, time-management and global use of technology by the organizations (Ribbers and
Waringa, 2015). Overall technological progress, with particular strong acceleration due to
COVID-19 pandemic impact, has expedited the progressive use of online technological tools
and significantly reduced face-to-face within the coaching landscape (Doolittle, 2023;
Terblanche, 2022). The recent COVID-19 pandemic crisis had a strong influence and

All authors contributed to the study conception and design. Material preparation, data collection and
analysis were performed by Lidia Plotkina. The first draft of the manuscript was written by Lidia
Plotkina and all authors commented on previous versions of the manuscript. All authors read and
Journal of Management
approved the final manuscript. Supervision was performed by Dr Subramaniam Sri Ramalu. Development
Funding: No funds, grants or other support was received. © Emerald Publishing Limited
0262-1711
Conflict of interest: The authors have no relevant financial or non-financial interests to disclose. DOI 10.1108/JMD-06-2024-0182
JMD accelerated the shift of coaching intervention into the online realm (Carnevale and Hatak,
2020; ICF, 2020). During the pandemic, in-person coaching reduced by 80%, and audio-video
means coaching increased by 74% (ICF, 2020). After the pandemic, the perspective of active
engagement in online coaching interaction with AI became more supported by users than
before (Schermuly et al., 2022). The expectations of expansive use of technologies for
coaching and belief that AI could simplify the process of coaching were reflected in the global
surveys.
Technological advancements in coaching are progressing rapidly, while comprehensive
experimental research has not yet fully developed. Contemporary trends in the field
increasingly incorporate e-coaching and AI-based coaching methodologies. E-coaching is a
format of coaching practice implemented via different technologies, so that it can be done
distantly between a coach practitioner and a coachee. It usually simplifies the procedure and
it is more cost-effective (Diller and Passmore, 2023). Meanwhile, AI Coaching is a
synchronous or asynchronous coaching using AI as a coach instead of a human coach
(Passmore and Tee, 2023a). Several narrow task coaching chatbots were created and tested
for efficiency and appreciation from the coachee’s side (Graßmann and Schermuly, 2021;
Movsumova et al., 2020; Terblanche et al., 2022a). Particular coaching tasks can already be
delegated to AI-based coaching chatbots, such as goal attainment, new health protocol
implementation support, education related coaching and more (Chew, 2022; Kocaballi et al.,
2019; Mai et al., 2021; Mitchell et al., 2021).
However, the principal shift in application of AI and machine learning models happened
with the launching of new multimodal large language models, such as generative pre-trained
transformer (GPT). These models allow to create sophisticated and complex dialogs
providing support and guidance in various fields (Carlbring et al., 2023; Lee et al., 2023).
Thus, GPT-based chatbots are different from the scripted and rule based chatbots that were
previously used. These tools can be identified by the following primary attributes: firstly,
they are designed for general-purpose applications rather than specialized ones; secondly,
they possess the capability to generate innovative language outputs that closely resemble
human communication; and thirdly, they provide a user-friendly interface that comprehends
and responds to natural language. Due to the promising preliminary results, the boom of the
AI-based coaching chatbot publications is expected. At the same time, chatbots usage raises
numerous ethical and privacy related questions, as well as the distribution of responsibility
and their efficiency (Cabrera et al., 2023). Meanwhile, the objective of this review is to analyze
existing approaches and AI-based coaching performance results, with the primary goal to
review AI-based coaching solutions and to compare the capabilities of AI tools with human
coaching.

2. Methodology
The purpose of this study is to conduct a systematic literature review on AI coaching. To
ensure impartiality, specific criteria were established at the beginning to determine which
articles should be included or excluded. Only high-quality and relevant studies from
reputable academic databases were searched and considered. A careful screening process
was implemented to ensure that the final selection of studies for the literature review aligns
with the research topic. The review is to provide the answers to the following questions:
Q1. What are the existing approaches or types of AI coaching?
Q2. What are the empirical evidences of AI coaching tools and chatbots implementation?
Q3. What advantages and disadvantages of AI coaching are derived from existing
studies compared with human coaching?
2.1 Selected review method Journal of
The systemic literature review aims to evaluate and compare the findings of different studies Management
and identify the gaps in the existing literature. To determine this, a systemic literature review Development
is conducted where studies were collected from academic databases like Scopus, Academia,
ResearchGate, Web of Science, and Google Scholar. These studies were considered with their
relevance to the current topic. To facilitate this, the titles and abstracts of the studies were
carefully examined. Systematic literature review based on the PRISMA diagram tool
consists of different stages and uses inclusion and exclusion criteria. Thus, the sources of the
studies have been identified, as well as the key words that are relevant for this research, the
timeline of this research, and actual adequacy of the studies in terms of their content. Precise
inclusion and exclusion criteria are listed below, alongside with the stages of identification,
selection, eligibility and final inclusion of the studies.
2.1.1 Inclusion criteria.
(1) Studies published between 2019 and 2023 were considered for this review as an
inclusion criterion.
(2) The studies included conceptual papers and empirical results on AI coaching
implementation.
2.1.2 Exclusion criteria.
(1) Studies that were not focused on coaching and contain other AI chatbots
implementation.
(2) The unavailability of the full text of the study. Studies with solely available abstracts
were excluded, due to the need to analyze the entire document to get full and deep
understanding of the research works.
(3) Thesis research studies were excluded in preference for the examination of peer-
reviewed journal submissions exclusively.

2.2 Data collection and processing methods

2.2.1 Stage I identification. All of the relevant to the topic studies were identified in the
academic database, such as Scopus, ResearchGate, Google Scholar, Academia, and Web of
Science. The keywords “coaching chatbot, AI coaching, e-coaching effectiveness, Artificial
Intelligence, GPT-4” were used for the literature search with the total number of final studies
of 339.
2.2.2 Stage II selection. Original studies were selected by eliminating duplicate studies
due to necessity as multiple academic databases were used in this study and hence the same
study can be found in more than one database. After the elimination of duplicate studies, the
number of remaining original studies was 175. These 175 studies were carefully analyzed
with regards to their titles and abstract for relevance. 84 more studies were excluded as they
were not relevant to the current research objective with post elimination total number of 91
remaining studies.
2.2.3 Stage III eligibility. The availability of academic literature can vary depending on the
access mode of the studies. For studies with restricted access, only the abstract is available.
However, the abstract alone is not sufficient for a full understanding of the study. Thus, 17
studies were additionally excluded with the remaining number of 74 studies, that were fully
available after this exclusion.
2.2.4 Stage IV inclusion. The remaining 74 studies were analyzed based on predetermined
inclusion and exclusion criteria for the study. The studies that do not fit in inclusion criteria
were excluded from the systematic literature review. Hence, 51 more studies were excluded
JMD based on that, to specify: not about coaching, but using chatbots - 20, about coaching, but no
AI component – 30, thesis work - 1. The final number of studies that were considered for this
systemic literature review is 23, which will be considered while analyzing the findings in the
literature. According to the most recent timeline, the strict criteria was selected and applied to
the studies in this review, all available and relevant literature on the implementation of AI
coaching chatbots in different contexts was included. Although 23 studies represent a
relatively small amount of evidence as the field of AI coaching is still emerging and growing.
Therefore, more theoretical and practical research on this topic is expected to appear in the
nearest future. Moreover, according to systematic literature review guidelines, there is no
proven minimum number of studies included (Kitchenham and Charters, 2007), and typically
SLRs could include from 10 and up to 50 articles. The former research works show the
existence of SLRs based, for instance, on 6 studies, and they managed to provide valuable
insights in the respective fields (Wang et al., 2023).
2.2.5 PRISMA flow diagram. PRISMA stands for “Preferred Reporting Items for Systemic
Review and Meta-Analysis” and can be characterized as a base arrangement of things that
can be utilized for meta-examination and foundational surveys. For the purpose of this study,
a PRISMA flow diagram (Figure 1) is created for a better understanding of how studies are
identified, included, and excluded from the systemic literature review. The PRISMA diagram
is essential for transparency, as it allows to see the method of studies’ selection and ensures
that the review process was systematic and replicable.

3. Results
AI systems could serve the coaching intervention in multiple ways (Strong and Terblanche,
2020). Machine Learning techniques allow to improve the process of coaching and could help
in coach-coachee matching. They could provide additional recommendations and resources
in between coaching sessions as well as the support for the goal attainment and coaching
advice (Khandelwal and Upadhyay, 2021a, b; Movsumova et al., 2020; Passmore and Tee,
2023a). The expansive scope of AI demonstrates that it offers more than just synchronous
coaching solutions. However, not all applications of AI in coaching are well-studied yet. In
this review, we focus on all the available studies about AI coaching and review their
performance, as well as current strengths and weaknesses.

3.1 AI-based coaching software and coach support

One of the useful AI-based software was suggested by Arakawa and Yakura (2020). A
specialized software was developed to analyze recorded coaching sessions. This software
provides a user-friendly interface that allows coaches to review sessions at an accelerated
pace and selectively access important parts of the conversation. Coaches can also take notes
and offer meta-reflection to the coachee. In summary, this tool simplifies the process of video
analysis and has shown to enhance the effectiveness of coaching sessions. In addition, a fully
automated system was developed to assist coach practitioners in identifying unconscious
behaviors exhibited by their coachees. This system utilizes an unsupervised anomaly
detection algorithm that analyzes multimodal behavior data, including posture and gaze. By
providing real-time feedback, this tool alerts coaches to relevant behavioral cues. The
algorithm generates informative cues that help coaches gain insight about the internal states
of the coachees. Furthermore, the use of an unsupervised machine learning algorithm
ensures that personal biases are avoided while an effective coaching support is provided
(Arakawa and Yakura, 2019, 2022).
Another AI-based coaching assistant was tested for its ability to provide help and
insights for the coaches and coachees (Movsumova et al., 2020). The Mentorbot appeared to
 Journal of
Identification of studies via databases Management
Development

Records identified from Google

Identification

Scholar, Scopus, ResearchGate,

Academia, Web of Science:
Databases (n = 5)
Registers (n = 328)

Records screened for duplicates Records excluded

(n = 175) (n =153)
Screening

Records assessed by titles and Studies excluded

abstracts (n = 84)
(n = 175)
Eligibility

Records assessed for eligibility Studies excluded:

(n =91) Not available full text (n = 17)

Studies excluded:
Included

Studies included in review Not about AI coaching (n =20)

(n =23) Not about chatbots (n =30) Figure 1.
Thesis (n = 1) PRISMA flow chart on
inclusion and
exclusion criteria
Source(s): Authors’ own creation/work

be helpful in suggesting the proper questions and provided deep and high-quality examples
of questions. The findings of the research indicate that an AI-based tool is more effective in
addressing new requests that are of high importance to the coachees in terms of their
willingness to take action and gain clarity. Whereas in terms of overall perception, the real
coach is seen as more useful, effective, and capable in facilitation of stress-management
focused sessions. The choice of the AI assisted tool versus human coach was more
preferrable by the coachees in cases of sensitive and confidential requests (Movsumova et al.,
2020). The increased reliance on machines might depend on the level of placing trust in the
privacy of the data exposed and shared during coaching sessions (Movsumova et al., 2020).
Additionally, the combination of coach and AI chatbot reveals that AI expands the coaches’
vision (Movsumova et al., 2020).
JMD 3.2 Performance of AI coaching chatbots
There is a noticeable advancement in the development of synchronous and asynchronous
coaching applications that offer a wide range of services (Passmore and Diller, 2024).
Different types of approaches could be utilized as performance algorithms of AI coaching
chatbot. For example, simple rule-based approach, scripted approach or reinforcement
learning data driven generation of coaching dialogs. Recently appeared Natural Language
Processing (NLP) based GPT models are the most sophisticated among them (Passmore and
Tee, 2023a). In the study conducted by Mitchell et al. (2022), the authors compared the
effectiveness of three approaches: scripted, rule-based and Reinforcement Learning (RL)
formulated interaction of coaching chatbot with a coachee about the health-related goals’
attainment like nutrition. The findings revealed that the data-driven RL chatbot performed
well in short conversations. Unexpectedly, the simplest scripted chatbot received higher
ratings in terms of quality, despite not consistently fulfilling its intended purpose. These
results underscore the conflicting nature between scripted and more intricate data-driven
approaches when it comes to chatbots in the healthcare domain. Therefore, there are cases
when the narrow tasks are executed more effectively with a straightforward scripted chatbot
rather than relying on complex yet inaccurate reinforcement learning responses and
prompts. The future direction of research might include the comparison of GPT based
responses and scripted responses in chatbots related to healthcare coaching.
The prominence of AI-driven healthcare coaching is on the rise, primarily attributed to its
cost-effectiveness and the round-the-clock availability of AI tools to assist coachees in pursue
of their goals. In the process of dealing with obesity and managing weight loss the support,
special recommendations and coaching are exceptionally helpful. Reinforcement learning
algorithms contributed to effective diabetes related health coaching intervention (Di et al.,
2022). To explore the factors that facilitate engagement of coachees in health coaching
interaction, the scoping review was conducted (Chew, 2022). The primary functions of AI
chatbots encompass the delivery of personalized recommendations, motivational messages,
gamification elements, and emotional support. According to the research, the speech with
appropriate colloquial tones in chatbots increased user engagement, due to convenience of
hands-free interaction, interactivity and expression of empathy in voice tones. Moreover,
additional strategies employed in text-based chatbots included emojis mimicking human
emotional expressions, the incorporation of positively framed vocabulary, referencing
credible information sources, personifying the chatbot, offering validation, and offer of real-
time, rapid, and valid recommendations (Chew, 2022). Moreover, ability to interact with a
chatbot on different platforms and devices promotes greater engagement. However, the
privacy concerns and user friendliness were the main constrains in using health AI coaching
chatbot for weight management. Concerning the “user burden”, researchers usually refer to
several constructs, such as difficulty to use a tool, the problems with social interaction,
physical, emotional and mental load while using a tool, time and finances that are spent while
using a tool (Suh et al., 2016).

3.3 AI coaching chatbots working alliance

Empathy and supporting attitude of the coach seemed to be important in real coaching
sessions (Cidral et al., 2023). However multiple studies of coaching chatbots show the
presence of coaching goal attainment with absence of empathetic behavior from a chatbot. A
comparison was made between the performance of a machine-driven chatbot and human-led
chatbot coaching. While the human-led conversations showed more empathy and support
from the coach, the coaches faced significant challenges and frustration when trying to adapt
to text-based virtual coaching. On the other hand, the scripted chatbot remained effective,
consistent, and encouraged the coachees to be more autonomous (Mitchell et al., 2021).
 Another study examined the impact of human coaching and AI coaching on goal Journal of
achievement over a period of 10 months (Terblanche et al., 2022a). The results revealed that Management
both human and AI coaches were significantly more successful in assisting coachees in Development
attaining their goals compared to control groups. Surprisingly, towards the end of the trial
period, the AI coach demonstrated comparable effectiveness to the human coach. As a result
of their findings, the authors suggested that AI could potentially replace human coaches who
employ simplistic, model-based coaching approaches. However, it is important to note that
AI performed poorly in terms of empathy and emotional intelligence, which are essential
qualities of the human coaches that make them irreplaceable at this moment (Terblanche
et al., 2022a).
At the same time, the attempt to enhance rule-based coaching chatbot with Reinforcement
Learning (RL) power in emotions recognition was done (Alazraki et al., 2021). The idea of
making a machine more empathetic and emotionally intelligent is based on the evidence of real
coaching sessions. This factor plays an essential role in working alliance establishing between
the coach and the coachee (Albizu et al., 2019). In the study, the authors introduced
methodology and computational approach for a digital coach designed to assist coachees in
implementing self-attachment therapy protocols. Their findings indicate that the platform
consistently receives higher ratings for empathy, user engagement, and overall usefulness
when compared to a basic rule-based framework (Alazraki et al., 2021). Overall, scripted
chatbots are effective for narrow and straightforward tasks, even though they barely could
create a working alliance with a coachee or build trustful and deep coach–coachee relationship.
The creation of working alliance appeared to be salient in coaching sessions and this
ability is low in current generation of chatbots (Graßmann et al., 2019; Mai et al., 2022). To
further investigate the working alliance in a chatbot–coachee relationships, chatbot’s self-
disclosure and information disclosure approaches were tested (Mai et al., 2021). The authors
used the StudiCoachBot that provides coaching with regard to exam-related anxiety. The
study focused on testing whether the chatbot’s self-disclosure would enhance students’ self-
disclosure and the effectiveness of coaching. The outcomes contradicted the primary
hypothesis, revealing that information disclosure was more effective than machine self-
disclosure for students. Essentially, comprehensive information about certain processes
appears sufficient, and the simulation of “human-like” issues by a machine is not necessary.
Whereas, overall experience of using chatbot was positively rated. It resulted in facilitation of
the reflection processes and students had the intention to use the bot again. The aspects
requiring further development were highlighted, including that chatbot lacks the flexibility
in answers and individualized approach (Mai et al., 2021). The follow-up study extended the
research about self-disclosure and information disclosure from chatbot with 201 participants
experiencing online coaching for study (Mai et al., 2022). They did not find statistically
significant differences with regards to presence or absence of any type of disclosure.
Consequently, the foundation of the working alliance between the coach chatbot and the
coachee is facilitated through the influence of additional factors that require further
investigation in forthcoming researches.
Another valuable investigation of working alliance between coaching chatbot and
coachee was conducted by Ellis-Brush (2021). The primary objective of the research was to
investigate the association between AI coaching and its potential for enhancing a coachee
self-resilience and working alliance initiation. Over the course of eight weeks, a group of 48
volunteers was granted access to WYSA, a mental well-being AI chatbot application. WYSA
is a downloadable software application chatbot accessible through smartphones. It utilizes
AI, natural language processing, and machine learning algorithms. It employs carefully
designed algorithms to generate conversations that mimic interactions with humans, all with
the aim of enhancing an individual’s self-resilience. WYSA achieves this by employing
evidence-based and validated tools and techniques, including cognitive behavioral therapy.
JMD The synthesis of both quantitative and qualitative data revealed important findings.
Although there was no substantial working alliance with the AI chatbot, the majority of
participants showed a notable improvement in self-resilience (Ellis-Brush, 2021). Drawing
upon multiple pieces of evidence, favorable outcomes could be discerned that are associated
with AI coaching even in the absence of a working alliance with a chatbot.

3.4 Factors affecting willingness to use coaching chatbots

The factors that influence the willingness to use coaching chatbots, and consequently impact
the effectiveness of AI coaching, are currently under investigation (Terblanche and Kidd,
2022). The chatbot was designed to focus on goal achievement, self-reflection, and non-
directive coaching. Terblanche and Kidd (2022) examined several factors that influence
users’ willingness to engage with AI coaching. These factors include users’ expectations of
the chatbot’s performance, the effort required to use it, the perceived impact of AI coaching,
potential risks, and the conditions that support its use. Among the five factors studied,
performance expectancy, facilitating conditions, and social influence were found to have a
statistically significant impact on users’ intent to use the chatbot. However, effort expectancy
and perceived risk did not show a significant influence.
To explore user experience and engagement in the coaching session, virtual
conversational bot was created based on Amazon’s ALexa platform (Kannampallil et al.,
2022). In the study, the authors developed a voice-based coach named Lumen, designed to
deliver an evidence-based problem-solving treatment program for depression and anxiety.
While participants emphasized the nonhuman aspect of the interaction, citing the absence of
variations in tone, emotion, instant feedback; almost all participants acknowledged the
potential benefits associated with Lumen’s accessibility. This allows individuals in need of
therapy to easily access a coach bot at any time. Participant highlighted the ways to reduce
the cognitive load and time pressure of interaction with coach chatbot so that sessions could
become more comfortable. The study gives insights about certain improvements that should
be taken into account while creating voice coaching tools and shows the potential of
developing the bond and understanding between the tool and the user (Kannampallil et al.,
2022). The authors suggest to use short and slow conversations, and allow to repeat and
pause the session. The primary challenges in achieving a natural conversational feel and
engagement revolve around the current limitations in language recognition and emotion
recognition systems, which, although promising, are still in the process of development.
Besides functions, the virtual appearance of the chatbot matters and influences its overall
success (Weber et al., 2021). The study was aimed to compare anthropomorphic chatbot with
less human-like appearing interface for goal attainment chatbot. The analysis revealed a
notable increase in satisfaction among users when interacting with the highly
anthropomorphic chatbot. Participants rated both the chatbot’s ability to build relationship
and the effectiveness of coaching more favorably compared to their interactions with the less
anthropomorphic chatbot. Consequently, it is evident that the level of anthropomorphism in a
chatbot used in online coaching sessions significantly influences its overall success. However,
more studies about chatbot persona are needed to carefully identify the amount of its influence.
Another essential variable to consider when evaluating the effectiveness of AI coaching is
the users’ age. Categories of users who grew up having mobile devices and are in constant
interaction with them are likely to perceive AI coaching more positively. For instance, a
health coaching AI chatbot was tested for its ability to support topics about depression and
anxiety in adolescents (Stephens et al., 2019). Adolescent users reported favorable
advancements towards their objectives 81% of the time. They exchanged thousands of
messages and reported high usefulness ratings while interacting with the chatbot. The study
showed that adolescents were actively engaged in the process and perceived it as a valuable
tool. Therefore, the potential for the devices’ use for coaching purposes is high for the Journal of
younger generation (Graßmann and Schermuly, 2021). Based on the results, we can Management
hypothesize that varying the interface and options of chatbots depending on the user’s age Development
could potentially increase performance and engagement (Hussain et al., 2018).

3.5 Generative pretrained transformer (GPT) AI coaching

Generative AI is a new promising field and these types of models could potentially overcome
several drawbacks that previous generations of machine learning chatbots had. The latest
emerging generative AI tools were tested in their abilities to create similar to real coaching
conversations (Passmore and Tee, 2023b). In the initial experiment ChatGPT had begun the
interaction by providing advice, although through various prompt manipulations,
particularly by prompting it to generate questions, a form of engagement resembling
coaching was managed to be established. It is worth noting that, even in this modified
context, the questions generated by ChatGPT were primarily multiple-choice questions,
often comprising sets of embedded answers or options, and none of them could be considered
suitable for use by a coach (Passmore and Tee, 2023b).
Meanwhile, in the second trial with ChatGPT-4 the performance had improved. The
authors provided chatbot with the particular prompts for facilitating personal responsibility
taking by the coachee and asking only one question at a time. In this case the coachee had
established the agenda of the coaching session, however chatbot took a leading role in the
conversation; this approach is usually unwanted from the real coaches. As advantages, the
assessor from the International Coaching Federation (ICF) acknowledged the presence of
empathetic responses, often initiated by reflecting upon or summarizing the coachee’s
statements. These responses were of supportive, positive, logical, and clear nature.
Nevertheless, GPT-4’s responses lacked in exploring coachee’s emotions, values and
individualized approach within the coaching dialog. The conversation was not fully able to
promote greater coachee’s responsibility taking and did not offer thought-provoking
questions that would have encouraged deeper introspection. Final evaluation concluded that
the transcript did not meet ICF Associate Certified Coach (ACC) standard. Moreover, in case
when coaching dialog contains a sensitive information about intention to suicide, the
response did not differ from regular conversation (Passmore and Tee, 2023b). This inability
to react to specific context makes the use of Chat GPT-4 for coaching unethical and
questionable. Even though the conversation with GPT-4 could express some empathy and
produce coaching-like questions, it is still in its early development stage. Therefore, the use of
generative AI in coaching sessions without supervision does not meet coaching standards
and coaching practitioners are invited to take actions to improve the technology.

3.6 Competencies of AI coaching

Based on current analysis, several capabilities of AI coaching are identified. In the initial
stage of AI coaching development, the list of competencies is not extensive. However, even in
this stage there are multiple ways to use it in different areas and for various purposes. The
summary of reviewed studies is presented in Table 1.
In summary, emerging tools and technologies have proven to be valuable in coaching
sessions, expanding the possibilities within the field, making the coaching more accessible
and widespread. Existing studies showed that some technologies enhance the coach’s
functionality, while others have the potential to replace basic coaching interventions. Proper
classification of AI coaching solutions can be achieved alongside the development of
emerging tools and studies in this growing field. Nonetheless, it is essential to acknowledge
that AI cannot replicate the full range of capabilities performed by a human coach. Therefore,
the adoption of new technologies is possible, though it is favorable to assess new solutions by
the professional coaches and provide justification of their effectiveness.
JMD Functionality Studies

Behavioral cues analysis, reflection acceleration Arakawa and Yakura (2019, 2020, 2022)
Provides AI-based analysis of coaching session video
Coach support Movsumova et al. (2020)
Gives appropriate hints and recommendations while coach is
at work
Goal attainment Terblanche et al. (2022a)
Collects updates, reminds of the goals, supports
Health protocols Chew (2022), Mitchell et al. (2021), Stephens
Provides necessary information about medical protocols, et al. (2019)
checks on health goals
Emotion recognition Alazraki et al. (2021)
Interprets emotions from phrases and emotional words,
expresses empathy
Study coaching, anxiety counseling Kannampallil et al. (2022), Mai et al. (2021),
Gives techniques and information, asks proper questions to 2022
eliminate anxiety
Mental health support Ellis-Brush (2021), Kannampallil et al.
Integrative support, psychoeducation, and interventions (2022), Stephens et al. (2019)
through brief conversations
Summarizing and speech understanding Passmore and Tee (2023a, b)
Empathy expression through speech processing
Reflection induction Passmore and Tee (2023b)
Table 1. Recognizes the speech and provides proper questions to
Functions of AI provoke thinking and problem solving
coaching tools Source(s): Authors’ own creation/work

3.7 Drawbacks and advantages of AI coaching and ethical considerations

From the principal perspective, International Coaching Federation (ICF) verified certain
competencies, that certified coaches should possess. Setting the foundation of trust and
rapport, co-creating the relationships, communicating effectively and facilitation of learning
and results are the main ones. Based on current evidences, we can assess the amount of
correlation between AI coaching chatbot and these competencies. AI coaching chatbot can
provide certain coaching competencies that align with ICF coaching standards to some
extent, but they are far away from passing the threshold of being a certified coach. Here are
some aspects in which AI coaching chatbots align with ICF coaching competencies. The
studies analyzed in the current review provide evidence and summary regarding this matter
in Table 2.
In summary, AI coaching chatbots can align with some ICF coaching competencies,
especially in areas like goal setting, active listening, and accountability. Overall, chatbots are
helpful in broadening of coaching practices with simple tasks. They could act more
consistent, persistent and straightforward than human coaches. However, they cannot
replicate the human coaching, particularly in terms of empathy, adaptability, and the ability
to address complex and unique individual requests. AI cannot provide long-term in-depth
work, that is a core pillar of coaching relationships. In most cases, AI chatbots still
significantly struggle to be coaches in a broad sense, their coaching approach lacks the
distribution of the responsibility and the coaching process remains vague. Ethical questions
about using GPT based AI coaching are still unsolved and this process should be supervised
by professional accredited coaches (Passmore and Tee, 2023b).
 Journal of
Advantages of AI chatbots Drawbacks of AI chatbots Management
Active listening and questioning Empathy and emotional support Development
AI chatbots can be programmed to actively listen to AI chatbots lack the ability to truly understand and
user input and ask open-ended questions to help empathize with a user’s emotions and experiences,
users explore their thoughts and feelings, which is a which is a crucial aspect of coaching (Terblanche
fundamental coaching skill (Alazraki et al., 2021; et al., 2022b). However, there is a positive progress in
Passmore and Tee, 2023b; 2023a) the emotion recognition and empathy translation
(Alazraki et al., 2021; Kannampallil et al., 2022;
Passmore and Tee, 2023b)
Goal setting Adaptability
Chatbots can assist users in setting and clarifying While AI can adapt to user input to some extent, it
their goals, which is an important aspect of coaching may not handle complex or unique situations as
(Chew, 2022; Terblanche et al., 2022b) effectively as a human coach (Kannampallil et al.,
2022; Passmore and Tee, 2023b; 2023a)
Feedback and reflection Complex problem solving
They can provide feedback, based on user responses AI chatbots may struggle with complex, nuanced
and encourage self-reflection, another key coaching issues that require in-depth exploration and problem-
component (Arakawa and Yakura, 2019, 2020, 2022; solving (Passmore and Tee, 2023b, 2023a)
Mitchell et al., 2022; Mitchell et al., 2021; Passmore
and Tee, 2023b)
Accountability Personalization
AI chatbots can remind users of their commitments Human coaches can provide highly personalized
and hold them accountable for their actions and coaching tailored to an individual’s unique needs and
progress towards their goals (Mitchell et al., 2022) circumstances, which AI chatbots may struggle to
replicate (Graßmann and Schermuly, 2021;
Kannampallil et al., 2022)
Information and resources Ethical and cultural sensitivity
They can provide relevant information and resources Human coaches can navigate sensitive ethical and
to support coaching process, such as articles, videos, cultural considerations that AI may not be equipped
or exercises (Chew, 2022; Movsumova et al., 2020) to handle (Passmore and Tee, 2023b)
Availability
AI chatbots can be available 24/7, allowing users to
access coaching support whenever they need it. This
is especially valuable in health goals or learning
goals, when constant support is needed along with Table 2.
some nudges and reminders to pursue new The competencies of
intervention (Chew, 2022; Kannampallil et al., 2022; AI coaching chatbots
Mai et al., 2021) that align with ICF
Source(s): Authors’ own creation/work certification

3.8 Theoretical and practical implications of AI coaching research

This review presents the comprehensive evidence on the practical applications of AI
coaching chatbots across various domains. Education, medical interventions,
psychotherapy, management and HR domains have implemented coaching chatbots (Di
et al., 2022; Fitzpatrick et al., 2017; Kannampallil et al., 2022; Khandelwal and Upadhyay,
2021b; Kocaballi et al., 2019; Maria et al., 2022; Mitchell et al., 2021; Passmore and Woodward,
2023; Stephens et al., 2019). Current review identified the potential of AI coaching chatbots
and provided evidence of their effectiveness. Gathered information is summarized in Table 1
and Table 2, which in detail delves into the functionality and competencies of these chatbots,
along with the advantages and disadvantages of their use. This review offers valuable
insights into the current theoretical landscape of coaching chatbots and serves as a useful
resource for practitioners. The findings of this article contribute to bridging the gap between
JMD theoretical knowledge about chatbot capabilities and their practical applications. While
highlighting the key aspects of AI coaching, it also underscores the areas which require
further exploration, particularly theoretical elements not covered in this study. Additionally,
this review delineates the actual capabilities of chatbots, enabling companies and
practitioners to make informed decisions about implementing these tools in their practice.
AI coaching chatbots are emerging as a potentially cost-effective solution that can make
coaching services more accessible. Studies have shown that the implementation of AI-driven
chatbots can result in significant cost savings, particularly in sectors like education,
healthcare, and corporate training, where individualized coaching is in high demand
(Terblanche et al., 2022a, b). The scalability of chatbots allows for broader application,
potentially leading to increased market penetration and enhanced customer engagement. In
educational settings, the findings from this review can inform the integration of coaching
chatbots into curriculum design and pedagogy. By utilizing chatbots as supplementary
teaching tools, educators can provide students with personalized support, thereby enhancing
learning outcomes and accommodating diverse learning styles (Mai et al., 2021). This
approach can be particularly valuable in large classes or online learning environments where
individualized attention from instructors is limited. The societal impact of coaching chatbots
is multifaceted, influencing public attitudes towards AI and potentially improving quality of
life. As chatbots become more prevalent in personal development, they can democratize
access to coaching services, making them available to individuals who might otherwise lack
the resources or opportunity to receive such support. This could lead to greater self-
improvement and well-being across diverse populations. However, it is important to
recognize the current limitations of AI coaching, which often involve delivering simplistic
approaches and short-term interventions. These interventions may fall short in establishing
a strong working alliance, offering individualized approaches, and providing the flexibility
needed for ethically sound guidance in complex situations (Passmore and Tee, 2023b, 2023a).
In some cases, narrow tasks may be executed more effectively by straightforward scripted
chatbots rather than relying on complex but potentially inaccurate responses generated by
reinforcement learning models (Mitchell et al., 2022; Mitchell et al., 2021). The advent of GPT-
based chatbots has the potential to significantly transform the coaching landscape, altering
how coaches deliver their services and how coachees engage with these tools. If security and
confidentiality of data can be guaranteed, and the effectiveness of GPT-based coaching
dialogs is validated, these AI interfaces could become a widely accepted and integral part of
the coaching process. This shift could lead to more accessible and scalable coaching
solutions, enabling individuals to benefit from personalized guidance without the limitations
of traditional in-person sessions. However, it is crucial to address the ethical and practical
challenges associated with this technology to ensure that its implementation is both
responsible and beneficial. As AI technology continues to evolve, there is a growing need for
regulatory frameworks that address the ethical implications of AI in sensitive areas like
education and mental health. Recent policy papers have highlighted the importance of
establishing guidelines for AI transparency, accountability, and user consent
(Shneiderman, 2020).

4. Discussion
AI coaching stands out as a potentially cost-effective solution that can render coaching
services in a more accessible form. However, it’s crucial to acknowledge that the current state
of AI coaching is limited by its tendency to deliver simplistic and short-term interactions.
These interventions often lack the establishment of a strong working alliance, individualized
approaches, and the flexibility required to provide ethically sound guidance in critical
situations (Passmore and Tee, 2023b, 2023a). Nonetheless, the outcomes of the review
underscore the suitability of integrating AI coaching into a myriad of contexts. AI coaching Journal of
applications have shown several promises in achieving a range of objectives, from goal Management
attainment and correction of medical conditions to psychological consultations. These Development
applications excel at suggesting pertinent self-reflective questions and, notably, can supply
information and education relevant to a given topic. Moreover, AI chatbots have
demonstrated their capacity to offer suitable recommendations, provide hints, convey a
sense of empathy, and interpret emotional expressions. Even though they show various
abilities, they cannot pass ICF coach certification requirements. They are definitely far from
substitution of the real coaching professional and being a full replacement is an unattainable
picture at this time. Additionally, the encouraging results achieved thus far are in tandem
with the rapid advancements in technology. However, it is essential to recognize that
scientific research and ethical considerations have not kept pace with these developments.
There exists an urgent need for the proper regulation of AI tool development to ensure that
interventions yield the anticipated results and that developers and supervisory coaches carry
responsibility for those results.
The direction for the future research could lay in exploring the distinctions in the
effectiveness of AI chatbots in different combinations and extent of human inclusion into
process. The ethical and cultural aspects are still an acute problem in creation of coaching
chatbot, so that extensive testing and adjustment of existing and emerging technology is
needed. Future research could explore a comparison between GPT-based responses and
scripted responses in chatbots. This comparison could provide deeper insights into the
strengths and limitations of different AI approaches in coaching.

5. Conclusion
This review discerned the capabilities and functionalities of contemporary AI coaching tools,
which have demonstrated remarkable proficiency in employing machine learning techniques
to analyze coaching sessions, facilitating reflective processes, and providing valuable
informational assistance, techniques and suggestions during coaching sessions. Their
capacities extend to substituting simplistic coaching programs, offering support in defining
and refining updates, plans, goals and even discerning the emotional phrases within
coaching interactions. Additionally, they are able to formulate relevant questions and serve
as reminders for coachees regarding their commitments and holding them accountable for
their actions.
However, it is essential to note that AI coaching tools fall short in replicating the
complexity of coach–coachee relationship and long-term real coaching interventions, often
unable to offer sophisticated and personalized approaches. Nevertheless, their capabilities
suffice for integration into executive coaching practices, complement existing approaches
and add extra support into coaching process.

References
Alazraki, L., Ghachem, A., Polydorou, N., Khosmood, F. and Edalat, A. (2021), “An empathetic AI
coach for self-attachment therapy”, Proceedings – 2021 IEEE 3rd International Conference on
Cognitive Machine Intelligence, CogMI 2021, pp. 78-87, doi: 10.1109/CogMI52975.2021.00019.
Albizu, E., Rekalde, I., Landeta, J. and Fern�andez-Ferr�ın, P. (2019), “Analysis of executive coaching
effectiveness: a study from the coachee perspective”, Cuadernos de Gesti�on, Vol. 19 No. 2,
pp. 33-52, doi: 10.5295/cdg.170876ea.
Arakawa, R. and Yakura, H. (2019), “REsCUE: a framework for REal-time feedback on behavioral
CUEs using multimodal anomaly detection”, Proceedings of the 2019 CHI Conference on
Human Factors in Computing Systems, pp. 1-13.
JMD Arakawa, R. and Yakura, H. (2020), “INWARD: a computer-supported tool for video-reflection
improves efficiency and effectiveness in executive coaching”, Proceedings of the 2020 CHI
Conference on Human Factors in Computing Systems, pp. 1-13, doi: 10.1145/3313831.3376703.
Arakawa, R. and Yakura, H. (2022), “Human-AI communication for human-human communication:
applying interpretable unsupervised anomaly detection to executive coaching”, ArXiv Preprint
ArXiv:2206.10987.
Cabrera, J., Loyola, M.S., Maga~
na, I. and Rojas, R. (2023), “Ethical dilemmas, mental health, artificial
intelligence, and llm-based chatbots”, International Work-Conference on Bioinformatics and
Biomedical Engineering, pp. 313-326, doi: 10.1007/978-3-031-34960-7_22.
Carlbring, P., Hadjistavropoulos, H., Kleiboer, A. and Andersson, G. (2023), “A new era in internet
interventions: the advent of Chat-GPT and AI-assisted therapist guidance”, Internet
Interventions, Vol. 32, 100621, doi: 10.1016/j.invent.2023.100621.
Carnevale, J. and Hatak, I. (2020), “Employee adjustment and well-being in the era of COVID-19:
implications for human resource management”, Journal of Business Research, Vol. 116,
pp. 183-187, doi: 10.1016/j.jbusres.2020.05.037.
Chew, H.S.J. (2022), “The use of artificial intelligence-based conversational agents (chatbots) for
weight loss: scoping review and practical recommendations”, JMIR Medical Informatics,
Vol. 10 No. 4, e32578, doi: 10.2196/32578.
Cidral, W., Berg, C.H. and Paulino, M.L. (2023), “Determinants of coaching success: a systematic
review”, International Journal of Productivity and Performance Management, Vol. 72 No. 3,
pp. 753-771, doi: 10.1108/IJPPM-07-2020-0367.
Di, S., Petch, J., Gerstein, H.C., Zhu, R. and Sherifali, D. (2022), “Optimizing health coaching for
patients with type 2 diabetes using machine learning: model development and validation
study”, JMIR Formative Research, Vol. 6 No. 9, e37838, doi: 10.2196/37838.
Diller, S.J. and Passmore, J. (2023), “Defining digital coaching: a qualitative inductive approach”,
Frontiers in Psychology, Vol. 14, 1148243, doi: 10.3389/fpsyg.2023.1148243.
Doolittle, J.S. (2023), “Virtual coaching is inevitable and effective”, 2022 Editorial Board, p. 24.
Ellis-Brush, K. (2021), “Augmenting coaching practice through digital methods”, International Journal
of Evidence Based Coaching and Mentoring, Vol. 15, pp. 187-197, doi: 10.24384/er2p-4857.
Fitzpatrick, K.K., Darcy, A. and Vierhile, M. (2017), “Delivering cognitive behavior therapy to young
adults with symptoms of depression and anxiety using a fully automated conversational agent
(woebot): a randomized controlled trial”, JMIR Mental Health, Vol. 4 No. 2, p. e19, doi: 10.2196/
mental.7785.
Graßmann, C. and Schermuly, C.C. (2021), “Coaching with artificial intelligence: concepts and
capabilities”, Human Resource Development Review, Vol. 20 No. 1, pp. 106-126, doi: 10.1177/
1534484320982891.
Graßmann, C., Schoelmerich, F. and Schermuly, C. (2019), “The relationship between working alliance
and client outcomes in coaching: a meta-analysis”, Human Relations, Vol. 73, doi: 10.1177/
0018726718819725.
Hussain, J., Ul Hassan, A., Muhammad Bilal, H.S., Ali, R., Afzal, M., Hussain, S., Bang, J., Banos, O.
and Lee, S. (2018), “Model-based adaptive user interface based on context and user experience
evaluation”, Journal on Multimodal User Interfaces, Vol. 12 No. 1, pp. 1-16, doi: 10.1007/s12193-
018-0258-2.
ICF (2020), COVID-19 and the Coaching Industry, International Coaching Federation, available at:
https://coachingfederation.org/app/uploads/2020/09/FINAL_ICF_GCS2020_COVIDStudy.pdf
Kannampallil, T., Ronneberg, C.R., Wittels, N.E., Kumar, V., Lv, N., Smyth, J.M., Gerber, B.S., Kringle,
E.A., Johnson, J.A., Yu, P., Steinman, L.E., Ajilore, O.A. and Ma, J. (2022), “Design and
formative evaluation of a virtual voice-based coach for problem-solving treatment:
observational study”, JMIR Formative Research, Vol. 6 No. 8, e38092, doi: 10.2196/38092.
Khandelwal, K. and Upadhyay, A.K. (2021a), “The advent of artificial intelligence-based coaching”, Journal of
Strategic HR Review, Vol. 20 No. 4, pp. 137-140, doi: 10.1108/SHR-03-2021-0013. Management
Khandelwal, K. and Upadhyay, A.K. (2021b), “Virtual reality interventions in developing and Development
managing human resources”, Human Resource Development International, Vol. 24 No. 2,
pp. 219-233, doi: 10.1080/13678868.2019.1569920.
Kitchenham, B. and Charters, S. (2007), “Guidelines for performing systematic literature reviews in
software engineering”, Technical Report EBSE 2007-001, Keele University and Durham
University Joint Report.
Kocaballi, A.B., Berkovsky, S., Quiroz, J.C., Laranjo, L., Tong, H.L., Rezazadegan, D., Briatore, A. and
Coiera, E. (2019), “The personalization of conversational agents in health care: systematic
review”, Journal of Medical Internet Research, Vol. 21 No. 11, e15360, doi: 10.2196/15360.
Lee, P., Goldberg, C. and Kohane, I. (2023), The AI Revolution in Medicine: GPT-4 and beyond,
Pearson, London.
Mai, V., Wolff, A., Richert, A. and Preusser, I. (2021), “Accompanying reflection processes by an AI-
based StudiCoachBot: a study on rapport building in human-machine coaching using self-
disclosure”, HCI International 2021-Late Breaking Papers: Cognition, Inclusion, Learning, and
Culture: 23rd HCI International Conference, HCII 2021, Virtual Event, July 24-29, 2021, Vol. 23,
pp. 439-457, Proceedings, doi: 10.1007/978-3-030-90328-2_29.
Mai, V., Neef, C. and Richert, A. (2022), “‘Clicking vs writing’—the impact of a chatbot’s interaction
method on the working alliance in AI-based coaching”, Coaching j Theorie and Praxis, Vol. 8
No. 1, pp. 15-31, doi: 10.1365/s40896-021-00063-3.
Maria, K., Drigas, A. and Skianis, C. (2022), “Chatbots as cognitive, educational, advisory and
coaching systems”, Vol. 30, pp. 109-126, doi: 10.47577/tssj.v30i1.6277, available at: www.
techniumscience.com
Mitchell, E.G., Maimone, R., Cassells, A., Tobin, J.N., Davidson, P., Smaldone, A.M. and Mamykina, L.
(2021), “Automated vs human health coaching”, Proceedings of the ACM on Human-Computer
Interaction, Vol. 5 CSCW1, pp. 1-37, doi: 10.1145/3449173.
Mitchell, E., Elhadad, N. and Mamykina, L. (2022), “Examining AI methods for micro-coaching
dialogs”, Conference on Human Factors in Computing Systems - Proceedings, Vol. 1, pp. 1-24,
doi: 10.1145/3491102.3501886.
Movsumova, E., Alexandrov, V., Rudenko, L., Aizen, V., Sidelnikova, S. and Voytko, M. (2020),
“Consciousness: effect of coaching process and specifics through AI usage”, E-Mentor, Vol. 86
No. 4, pp. 79-86, doi: 10.15219/em86.1485.
Passmore, J. and Diller, S.J. (2024), “Defining digital and AI coaching”, in The Digital and AI Coaches’
Handbook: The Complete Guide to the Use of Online, AI, and Technology in Coaching, pp. 21-33,
doi: 10.4324/9781003383741-3.
Passmore, J. and Tee, D. (2023a), “Can chatbots replace human coaches? Issues and dilemmas for the
coaching profession, coaching clients and for organisations”, The Coaching Psychologist,
Vol. 19 No. 1, pp. 47-54, doi: 10.53841/bpstcp.2023.19.1.47.
Passmore, J. and Tee, D. (2023b), “The library of Babel: assessing the powers of artificial intelligence
in knowledge synthesis, learning and development and coaching”, Journal of Work-Applied
Management, Vol. 16 No. 1, pp. 4-18, doi: 10.1108/JWAM-06-2023-0057.
Passmore, J. and Woodward, W. (2023), “Coaching education: wake up to the new digital and AI
coaching revolution”, International Coaching Psychology Review, Vol. 18 No. 1, pp. 58-72, doi:
10.53841/bpsicpr.2023.18.1.58.
Ribbers, A. and Waringa, A. (2015), E-coaching: Theory and Practice for a New Online Approach to
Coaching, Routledge.
Schermuly, C.C., Graßmann, C., Ackermann, S. and Wegener, R. (2022), “The future of workplace
coaching – an explorative Delphi study”, Coaching: An International Journal of Theory,
Research and Practice, Vol. 15 No. 2, pp. 244-263, doi: 10.1080/17521882.2021.2014542.
JMD Shneiderman, B. (2020), “Bridging the gap between ethics and practice: guidelines for reliable, safe,
and trustworthy human-centered AI systems”, ACM Transactions on Interactive Intelligent
Systems (TiiS), Vol. 10 No. 4, pp. 1-31, doi: 10.1145/3419764.
Stephens, T.N., Joerin, A., Rauws, M. and Werk, L.N. (2019), “Feasibility of pediatric obesity and
prediabetes treatment support through Tess, the AI behavioral coaching chatbot”,
Translational Behavioral Medicine, Vol. 9 No. 3, pp. 440-447, doi: 10.1093/tbm/ibz043.
Strong, N. and Terblanche, N. (2020), “Chatbots as an instance of an artificial intelligence coach”, in
Coaching im digitalen Wandel, Vandenhoeck & Ruprecht, pp. 51-62, doi: 10.13109/
9783666407420.51.
Suh, H., Shahriaree, N., Hekler, E.B. and Kientz, J.A. (2016), “Developing and validating the user
burden scale: a tool for assessing user burden in computing systems”, Conference on Human
Factors in Computing Systems - Proceedings, pp. 3988-3999, doi: 10.1145/2858036.2858448.
Terblanche, N. (2022), “Coaching during a crisis: organizational coaches’ praxis adaptation during the
initial stages of the COVID-19 pandemic”, Human Resource Development Quarterly, Vol. 34
No. 3, pp. 309-328, doi: 10.1002/hrdq.21490.
Terblanche, N. and Kidd, M. (2022), “Adoption factors and moderating effects of age and gender that
influence the intention to use a non-directive reflective coaching chatbot”, Sage Open, Vol. 12
No. 2, doi: 10.1177/21582440221096136.
Terblanche, N., Molyn, J., De Haan, E. and Nilsson, V.O. (2022a), “Coaching at scale: investigating the
efficacy of artificial intelligence coaching”, International Journal of Evidence Based Coaching
and Mentoring, Vol. 20 No. 2, pp. 20-36, doi: 10.24384/5cgf-ab69.
Terblanche, N., Molyn, J., de Haan, E. and Nilsson, V.O. (2022b), “Comparing artificial intelligence and
human coaching goal attainment efficacy”, PLoS One, Vol. 17 No. 6, e0270255, doi: 10.1371/
journal.pone.0270255.
Wang, X., Edison, H., Khanna, D. and Rafiq, U. (2023), “How many papers should you review? A
research synthesis of systematic literature reviews in software engineering”, 2023 ACM/IEEE
International Symposium on Empirical Software Engineering and Measurement (ESEM),
Vol. 14, pp. 1-6, doi: 10.1109/esem56168.2023.10304863.
Weber, U., Loemker, M. and Moskaliuk, J. (2021), “The human touch: the impact of
anthropomorphism in chatbots on the perceived success of solution focused coaching”,
Management Revue, Vol. 32 No. 4, pp. 385-407, doi: 10.5771/0935-9915-2021-4-385.

Corresponding author
Lidia Plotkina can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]

View publication stats