ICF Coaching Platform Coalition: Shaping ethics on coaching platforms
The ICF Coaching Platform Coalition brings together experts to explore quality and ethics in the future of coaching.
Wisdom Weaver
ICF Coaching Platform Coalition explores ethics and standards associated with coaching platforms
The ICF Coaching Platform Coalition is a collaborative initiative between the International Coaching Federation and coaching platform leaders. As coaching on platforms and artificial intelligence (AI) disrupt the coaching industry, the coalition is working to define standards and operating procedures for coaching engagements. The aim is to ensure a safe and impactful coaching experience for coaches and clients through platforms.
Recognizing the growing demand for coaching and the emergence of AI tools in coaching spaces, the coalition is working to increase digital literacy and engage coaches in co-creating AI solutions. The group has hosted ICF Research Webinars to help coaches learn about AI and ask questions from AI coaching experts. As part of their efforts, they have established a special working group for Artificial Intelligence Coaching Standards, aiming to provide guidance on the development, application, and use of safe, ethical, and beneficial AI coaching services. The group has outlined the following priorities:
- To guide the convergence of AI technology and coaching
- To provide a set of principles, capabilities, and requirements to enhance the quality of AI coaching
- To ensure AI coaching is applied in productive and ethical ways
Anticipating AI pitfalls to ensure safe and unbiased AI coaching
With the explosion of generative large language model AI and AI coaching chatbots, the industry is quickly working to understand the potential benefits and challenges of AI coaching. The Coaching Platform Coalition notes that data security, privacy, safety, bias, and client autonomy can all be jeopardized through faulty technology. Through intentional design, testing, and use developers can create AI algorithms and train and test them in ways that reduce human bias, support safety, and provide more opportunities for clients to connect with coaching services. AI coaching applications that utilize machine learning will have safeguards in place that will prevent the introduction of bias from usage data.
“AI has blossomed tremendously in the last six months with the emergence of generative large language model AI and coaching chatbots. We currently have effective rules-based AI coaching platforms, and in the next year, we’ll see even more effective AI coaching platforms using generative large language model technology.”
Reflecting on developments in coaching and technology, Wisdom Weaver Joel DiGirolamo highlights that “generative large language model and deep learning AI systems such as ChatGPT and GPT-4 will transform AI coaching. When you tune those algorithms to coaching, they can be powerful systems with high impact.” As the Vice President of Research and Data Science at ICF, Joel reflects that collaboration among researchers and coaching platform leaders is advancing coaching science and expanding coaching impact.
Read the interview with Wisdom Weaver Joel DiGirolamo to learn more about how AI is transforming the future of coaching:
Interview with Joel DiGirolamo on ensuring quality and avoiding bias in the future of technology and coaching
What trends are you seeing with the confluence of technology and coaching?
“People have written for years about coaching for sustained change. Those days are gone. We need to be coaching for adaptability and resilience because things change so quickly now. Look at the pandemic: in two or three months, remote work created a new paradigm for management and teams. AI is similar. When we come out the other side, jobs will be displaced—absolutely. But will our productivity go up as a global society? I hope so. And there will be new jobs that will be available that incorporate this new technology. If you study history, this is what you will find.
The use of AI coaching platforms is only scratching the surface right now. And with analysis tools, there will be more and more ways to observe the fluidity of conversation or for coaches to learn nuances of their coaching. Then standalone coaching systems will proliferate more and more so that coaching as a modality can be introduced to more people. There will be a profound shift in the profession, where technology will help us take the idea or the concept of coaching and make it more concrete. So, when we say, ‘the use of coaching skills,’ those will be better defined, and that can enhance the profession.”
How is research supporting the development of AI coaching technologies?
“Everybody is learning together. Early on, therapy AI platforms were ahead of coaching, and there were not any coaching platforms at the time. Now, there are many AI coaching platforms out there. When we look at coaching research, we draw on therapy research, especially in supervision and exploring how coaching works. There are similarities. Right now, coaching research is leading the way. People are researching and investing money into how to take Open AI code and make more effective coaching systems. And looking at how they can assess coaching or look for poignant moments in the coaching session, that can lead to significant coaching productivity.”
How can emerging technologies in coaching support improved productivity and enhanced quality?
“AI coaching platforms can help in many ways. We already have one system that will listen and watch the coaching session, and it will evaluate how much the coach is speaking versus how much the client is speaking and can also identify peak moments in the dialog. It will help coaches who are engaging in ongoing professional development to identify places in the sessions where they could have done better or bring awareness to their personal coaching style. Similarly, coaching education programs can use these as tools as adjuncts to their human training or education.
This is currently an asynchronous approach, but there will likely be synchronous or real-time applications of AI where the AI system could be bringing up prompts to the coach. It could recognize certain patterns with clients or list questions the coach could ask clients that the coach hadn’t thought of. As humans, we are not very good at multitasking, but computers are really good at it. They could help us when we are coaching and trying to listen and be present to support options and decisions on where to go next.”
What challenges should coaches and software developers keep in mind when using AI in coaching?
“There is the problem with bias in that a system can only respond in ways that it’s been trained and with information it has been exposed to. AI systems with machine learning have the capability to add or enhance the quality of the system as it is used. Unfortunately, the data being fed in as it is used can introduce additional bias because that’s additional information that the system is training itself on. Depending on who is using the system and the clients they’re working with, there can be a large subset of populations using that system, and the system doesn’t learn to work with other subsets of the population. So that’s part of the bias problem.
Bias can also be an issue if coaches who are using these systems are using certain techniques. If coaches are using System A, then when a second group of coaches is using System B, the AI might not respond well to them. So, the machine becomes biased toward System A.”
How can the Coaching Platform Coalition and software developers work to ensure safe coaching products?
“The systems can be trained to make that differentiation between coaching and other fields. And there are words or phrases the system can key on. A lot of that has to do with the prompting and training and the tuning of systems. We have a lot of smart people who are working on these systems who will be able to tune them to separate coaching and therapy or guide the system back to certain topics.
The big concern is the huge problem of misinformation or poor-quality of information. With a bad actor, generative systems could be creating false information or information that is hateful and setting up websites that would spread this misinformation. Then, when other systems are crawling all those websites, it can create a compounding problem with misinformation. To address this, there need to be systems that can guarantee the provenance and quality of data to ensure trust.”
What one thing would you convey to coaches about how they can engage with technology in the future of coaching?
“Coaches need to get comfortable with technology and try these systems. You can try ChatGPT for free or test the standalone AI coaching systems. How can you offer that to your clients between your human-human sessions? And then, could you use those platforms as an adjunct to human-human coaching so that between coaching sessions, your client is engaging with a coaching bot? Then you can have more productive sessions, dig deeper into some of the issues, and that’s where humans come into play.
Change is inevitable. Unfortunately, it can cause a lot of angst. The universe is constantly moving and flowing, so getting on board with technology and becoming comfortable with it is important. Much like coaches talk to their clients about moving outside of their comfort zone, maybe they are going to have to get outside of their comfort zone too. There will be dramatic shifts in the profession. And when you look overall at what it’s going to do for humanity, these shifts could be a powerful force for good. Ask yourself, what can you do individually today as a coach to be a part of that progress, to keep the profession moving forward, and to spread the use of coaching skills for enhanced dialogue in all our cultures?”
