AI-Based Coaching: 5 Considerations for HR
Artificial intelligence has taken the world by storm. The release of Chat GPT and other applied AI models has brought opportunities for using AI HR into the spotlight again. One of the practical applications of AI in HR is AI-based coaching. In this article, we discuss how AI is currently used in coaching, the benefits and limitations of AI related to coaching, and five guidelines for HR to consider when adopting AI coaching.
Contents
What is AI-based coaching?
Types of AI-based coaching
What are the benefits and limitations of AI-based coaching?
5 HR considerations for implementing AI-based coaching
What is AI-based coaching?
AI-based coaching is the use of artificial intelligence to support, enable, augment, complement, or take ownership of the coaching relationship. To get slightly more technical, researchers define AI in coaching as the “machine-assisted, systematic process of helping clients set professional goals and construct solutions to achieve them”.
As practical applications of AI in HR continue growing, so do challenges concerning bias, lack of transparency, and ethics. A case in point refers to New York City incorporating new legislation to regulate the use of AI within hiring practices. To overcome these challenges, we need to become familiar with the different uses of AI in HR to make the most of it in an effective and ethical way.
In our video series HR Dialogues, we discussed the future of coaching with Dr. Glenn Wallis from a business coaching company Exigence. See the full interview below:
Types of AI-based coaching
There are three different types of AI-based coaching based on the role AI plays in the coaching relationship. Let’s take a closer look at these.
AI-supported coaching
What is it? AI-supported coaching means that coach uses AI to gain insight into coaching needs and provide recommendations to inform the coaching relationship.
Example: A coach utilizes AI-based assessments and asks the coachee to engage with AI-driven tools to provide information to the coach. This data improves the quality and effectiveness of the coaching process.
Vendor example: Coachhub
AI-augmented coaching
What is it? AI-augmented coaching refers to the practice where coachees engage with AI-based tools between formal coaching engagements with a human coach. AI plays the role of a guide to promote further development and helps the coaching process continue beyond face-to-face sessions.
Example: A coachee engages with their coach once a month for a 90-minute session. In between, they access an AI-driven journey of developmental tasks that pushes content to the coachee based on assessments and priorities and tasks that their coach identified. The coach can also access these insights and uses the AI-driven journey to complement the human-led sessions.
Vendor example: Exigence, Centrical
AI-as-the-coach
What is it? AI-as-the-coach is a practice where AI is the coach and individuals only engage with AI. There is no or limited interaction with a human coach, and the coaching relationship sits between AI and the coachee.
Example: A coachee uses a recognized AI coaching system to improve their current level of self-awareness.
We believe that “AI-as-the-coach” approaches will become more commonplace in the future with the development of AI tools with more advanced language generation capabilities. Still, the research on the effectiveness and ethics of “AI-as-the-coach” without human oversight is not yet conclusive.
Given the shortage of coaches and the increasing need for coaching services that are affordable and scalable, coaching will likely become more democratized in the future.
Vendor example: Coach Vici, QTrobot, Replika, Misty II Robot
What are the benefits and limitations of AI-based coaching?
All three applications of AI coaching have different benefits and limitations. It is essential to understand these when incorporating AI into your coaching practices.
Benefits Limitations AI-supported coaching – Improved quality of coaching sessions
– Improved effectiveness of the coach– Limited benefit beyond the coaching session for the coachee
– Not scalableAI-augmented coaching – Improved quality of coaching experience
– Strong data-driven and evidence-based approach
– Continuous coaching in “the nature of work and life”
– More scalable than traditional coaching models– Can be costly and is usually subscription-based
– Requires clear role boundaries between the coach, AI, and the coachee
– Potential risk of breeding dependency over timeAI-as-the-coach – Scalability
– Accessibility– Limited research to demonstrate safe use and impact
– No oversight from a certified professional
– Cannot yet deal with complex challenges and contextual matters
When used effectively and in the right way, organizations can gain significant value from introducing AI-based coaching practices. But where do you start? And what guidelines can help you consider what is the right fit for your organization?
5 HR considerations for implementing AI-based coaching
We propose five considerations for HR to identify whether AI-based coaching is a good fit for your organization and how to approach it ethically and responsibly.
1: Clearly define the purpose of coaching and the scope
The first guideline is to be clear on the purpose of the coaching process, its scope, and its boundaries. Organizations can benefit from using AI for coaching objectives related to leadership development, personal growth, and performance, to mention a few examples.
AI is unsuitable for dealing with serious mental health challenges and should not, in any event, form part of your practice. Certified healthcare practitioners should always design and implement clinical matters interventions. While a certified healthcare professional could potentially benefit from AI support, AI should not be used for direct engagement with the coachee.
2: Consider the complexity and potential risk for harm
Once you’ve identified the purpose of the coaching engagement, evaluate the complexity and potential risk for harm.
Simply put, this refers to the potential risk to the coachee if something goes wrong with the AI interaction. A registered coach works from the perspective of “Do no harm”; yet this is not necessarily true for AI. For example, dealing with mental wellbeing challenges poses a higher risk of potential harm than managing a career transition.
Using these two criteria, the figure below helps you understand how and where to apply the three approaches.
- AI-as-the-coach is applicable in low-complexity and low-risk situations as it struggles to understand context and nuance. It is, however, effective in dealing with “narrow” and predictable situations. We see great applications of this practice for general developmental purposes for entry- or mid-level individuals.
- AI-augmented coaching is suitable for dealing with low/medium complexity and low/medium risk situations, given the oversight of a professional that complements the engagements with AI. This can be very useful for mid to senior-level general development.
- AI-supported coaching is best for dealing with high-complexity and high-risk situations, as the traditional oversight and principles will guide the coach/coachee relationship. The role of AI here is based on improving the quality of the coaching relationship through data and evidence-based practice. However, the limitation is that this approach is time-intensive and not scalable. Senior/executive-level coaching practices could benefit most from this approach.
3: Create the appropriate awareness, oversight, and controls
When using AI-based coaching approaches, it is crucial that all parties clearly understand how AI will be used as well as its limitations.
For example, the coachee needs to be aware of what data shared with the AI will be visible to the coach, how the data will be used to inform the coaching relationship, and the purpose of their interaction with the AI tool. Similarly, coachees need to understand how AI complements the relationship with the human coach and that the principles of maintaining clear boundaries, confidentiality, safety, and avoiding harm still apply.
In addition to clearly defining roles, it is also essential to implement appropriate oversight and controls.
For instance, if an individual being coached by “AI-as-the-coach” expresses suicidal thoughts or discusses topics that require the expertise of a mental health professional, there has to be a mechanism in place that escalates this matter for intervention. This is similar to what organizations such as Meta use to identify keyword content that could require the attention of a healthcare practitioner on their platform.
4: Ensure using a recognized coaching model with transparency on what data is used
When deciding on an approach, you need to assess the credibility of the coaching models and frameworks. Considering the quality of the data that will be used to inform the coaching process is also key. Incorporating AI into coaching requires an understanding of the coaching model that the AI is drawing from to produce its responses.
When working with vendors, ensure that they are able to supply sufficient evidence of the coaching framework and approach used. They should also explain how these are integrated into the sources that the AI utilizes to learn and respond.
Similarly, it is important to understand how AI utilization is validated over time. For example, how frequently does the vendor assess the quality of the AI responses?
In other words, to make an informed decision, you need to have confidence in the coaching approach and understand how AI works within its context. By assessing the credibility of models and frameworks, the quality of data, and the AI validation process, you can ensure that your coaching program is effective, efficient, and reliable.
5: Treat AI-based coaching as a complement to your other people development activities
Lastly, AI-based coaching needs to be incorporated into the organization’s larger development philosophy and activities. This sounds logical, but it becomes a lot trickier when we start evaluating the existing content and development paths within the organization.
At a minimum, you should be able to highlight how and where coaching is used in conjunction with other developmental activities. At best, the actual content should have a similar theoretical foundation to ensure consistency and alignment with your internal leadership and competency models.
As AI gets more advanced, this will become easier. However, for now, it is key to help coachees understand how AI-based coaching fits into the broader employee development plans and activities. When using AI-supported coaching, it is a good idea to brief the coach on the internal philosophies and how they can leverage those as part of the practice.
A final word
As we move forward, AI will start to play a more significant role in democratizing, scaling, and making coaching accessible to more people around the world.
By adopting a responsible approach to AI-based coaching, using it appropriately, and adhering to set guidelines when integrating it into wider people development practices, HR can capitalize on a valuable opportunity to bolster its impact on professional and career development.
Weekly update
Stay up-to-date with the latest news, trends, and resources in HR
Learn more
Related articles
Are you ready for the future of HR?
Learn modern and relevant HR skills, online