Introduction
The rapid development of generative artificial intelligence poses fundamental questions for coaching psychology regarding human agency. This article analyzes whether AI is merely a tool for increasing efficiency or if it can genuinely support human growth.
Readers will discover the risks associated with the automation of cognitive processes and learn how to design hybrid systems. The text is based on an analysis of theses from The Coaching Psychology Handbook, outlining a path toward maintaining autonomy in a world governed by algorithms.
AI in Coaching: Between Task Efficiency and Competence Development
The primary ethical and cognitive challenge is the discrepancy between the outcome of a task and the actual growth of the user's competence. Jonathan Passmore warns against the phenomenon of cognitive offloading—the shifting of mental operations to AI—which can lead to the atrophy of independent critical reasoning.
A prime example is automated goal setting. A system may create a perfect action plan, but if the user adopts it unreflectively, they will fail to develop strategic thinking skills. The author emphasizes that success should not be measured by the quality of the AI's response, but by the permanence of the change within the human being.
Distributed Cognition and the Extended Mind as Foundations for Digital Support
The philosophical foundations of using AI are based on the concept of distributed cognition and the theory of the extended mind. These theories posit that cognitive processes do not end at the biological boundary of the skull but encompass interactions with tools and the environment.
In this view, AI is not merely an external program but an active element of the cognitive system. This allows technology to be treated as an extension of human capabilities, provided the user retains the ability to verify results and understand the limits of the algorithm's reliability.
The Risk of Cognitive Offloading and Loss of Independence
Overreliance on AI can lead to automation bias, or the uncritical trust in a system's suggestions. Jonathan Passmore points out that the automation of reflection—the interpretation of values and priorities—is far more dangerous than auxiliary automation (e.g., calendar organization).
To prevent dependency, the author proposes a strategy known as scaffolding. This involves gradual support: AI initially assists with difficult tasks, but over time reduces its help to force independence. The goal is not to maximize app usage, but for the human to reach a level of competence that allows them to eventually dispense with the technology.
Summary
The true measure of maturity in digital coaching lies not in how many tasks the machine takes over for us, but in how much more autonomous we become because of its presence. Jonathan Passmore reminds us that while AI can support goals, it cannot replace human responsibility and ethical judgment.
Ultimately, the greatest risk is not that AI will replace the coach, but that humans will stop needing their own thinking to achieve their goals. The key to success remains digital agency—the conscious and critical steering of technology.
Frequently Asked Questions
What is the main ethical and cognitive problem associated with the use of AI in the coaching process?
The main problem is the difference between the automation of cognitive activities and the actual development of human competencies, which may lead to a limitation in the independent practice of skills. There is a risk that the user will unreflectively accept AI suggestions instead of developing autonomy in decision-making and self-regulation.
What philosophical foundations allow artificial intelligence to be treated as an extension of human cognitive capabilities, rather than just an external tool?
These foundations consist of the concepts of distributed cognition, according to which cognitive processes occur in systems comprising both humans and tools, and the extended mind, which assumes that external elements can be functionally integrated into the cognitive system. In this view, AI is not merely a passive tool, but an active element of the system that, through feedback, co-creates the way problems are understood and information is transformed.
How can the use of AI in cognitive processes affect a human's ability to think independently?
The use of AI can support critical reasoning, but it can also replace the independent formulation of premises and the assessment of their reliability. Overreliance on tools and automation bias may lead to reduced engagement in thought processes and the surrender of the role of arbiter—determining important problems and solutions—to AI.
What are the risks associated with the automation of thought processes in coaching, and how do the AI technologies used in it differ from one another?
The main risk is the paradox of assisted self-regulation, where excessive delegation of tasks to AI may limit the development of skills for independent problem-solving. AI technologies differ in their level of sophistication: from rule-based systems and chatbots executing programmed sequences, through machine learning tools for data analysis, to generative models capable of open dialogue.
Does scientific research confirm that AI can replace a human coach in achieving goals?
Research indicates that algorithmic systems can be effective in supporting the achievement of specific goals, organizing actions, and monitoring progress, without showing clear inferiority compared to human coaching. However, this does not mean full replaceability of professionals, as AI may struggle with ethical issues, value conflicts, or complex interpersonal relationships.
Do the successes of chatbots in psychological therapy mean that AI is equally effective and safe in coaching?
No, the success of specialized clinical systems does not automatically mean that the effectiveness of AI can be directly translated to coaching. Professional implementation of technology requires an assessment of the transferability of research results to a specific application environment and consideration of continuous changes in generative models.
Can a chatbot be truly empathetic, and what are the risks associated with treating AI as a human coach?
A chatbot does not possess affective empathy in the human sense; it merely generates utterances consistent with cultural patterns of empathy. Treating AI as a coach carries the risk of over-attributing competence and responsibility to the system and may lead to emotional dependency or the limitation of real social contacts.
Can AI in coaching build a genuine relationship with a client and correctly diagnose their emotional state?
An AI system can support goal organization and evoke a subjective sense of closeness, but it does not build a true relationship based on mutual accountability and moral commitment. Furthermore, AI linguistic analysis is not identical to a psychological diagnosis; the system can identify emotional phrasing, but it lacks the premises to reliably determine the user's actual emotional state.
What risks are associated with the automated analysis of emotions and communication in coaching, and how can AI be safely implemented in a hybrid model?
The main risks include the possibility of misinterpreting emotions due to cultural or linguistic differences or neurodivergence, as well as legal issues regarding the ban on emotion recognition in work and education. Safe implementation of AI in a hybrid model requires ensuring meaningful human oversight, a transparent contract, a competent professional, and treating algorithmic results as material for reflection rather than indisputable knowledge.
How should AI in coaching be designed to support the development of autonomy instead of creating dependence on technology?
To support autonomy instead of dependence, AI should be designed using a scaffolding model, meaning it gradually reduces the scope of assistance as the user's competence grows. It is crucial that the system is chosen and controlled by the client, and that the success of the intervention is defined by achieving a skill level that allows for reduced use of the tool.
In what ways can AI systems in coaching threaten user autonomy and become tools for manipulation or surveillance?
AI systems can threaten autonomy when they become tools for institutional monitoring and surveillance by classifying behaviors and setting efficiency norms. The risk of manipulation arises from dynamically adjusting the choice architecture to the user and employing engagement-reinforcing mechanisms that may serve the service provider's goals rather than the human's autonomous goals.
What legal regulations and data protection principles must be met when implementing AI in coaching?
Implementing AI in coaching requires compliance with GDPR regarding the protection of personal data (especially sensitive data) and the AI Act, which imposes transparency and risk management obligations. Users must be informed about their interaction with an AI system, the principle of privacy by design must be applied, and responsibility for decisions made by the technology must be precisely defined.
How can AI support neurodivergent individuals, and what risks are associated with adapting technology to their needs?
AI supports neurodivergent people through the personalization of interaction pace, the way information is presented, and the removal of environmental barriers. However, a risk exists in profiling—automatically drawing conclusions about a user's diagnosis or competencies—as well as the risk of making technology a prerequisite for participating in development programs.
What specific principles and criteria should guide the design of AI systems in coaching to ensure they are safe and ethical?
The design of AI systems in coaching should be based on the distinction between technological capability, psychological justification, and the ethical authorization to intervene. Systems must be transparent regarding their limitations and the interests of the operators, and their value should be measured by the extent to which they expand the user's actual capabilities. In a layered model, digital functions should handle informational and organizational matters, while complex problems must require the involvement of a professional.
What competency standards and research criteria are necessary for AI in coaching to support genuine human development rather than just automating tasks?
It is essential to develop coaches' digital competencies in understanding AI limitations, privacy protection, and the ability to distinguish algorithmic interpretation from psychological knowledge, as well as introducing supervision over the use of these tools. Research criteria should go beyond user satisfaction, focusing on the sustainability of results, competence transfer mechanisms, and the human's ability to make independent decisions after the process is completed.
What criteria should be used to assess whether AI in coaching actually supports human development rather than just efficiency?
The assessment of human development support should be based on criteria of autonomy, moral reflection, and the real expansion of the range of choice. Key factors include transparency of purpose, the possibility of refusal, privacy protection, and avoiding the hidden shaping of the user's beliefs.
What is digital agency and why is it crucial for maintaining the autonomy of a human using AI in coaching?
Digital agency is the ability to understand the role of technology in decision-making processes, assess its reliability, and control the scope of functions delegated to it. It is crucial for maintaining autonomy because it allows one to avoid unreflective submission to AI systems and enables conscious decision-making and taking full responsibility for those decisions.
What is the final conclusion regarding the role of AI in coaching and how should the success of this technology be measured?
The role of AI in coaching should be to support human autonomy and competence, while maintaining human responsibility for goals and moral judgment. The success of this technology is not measured by the number of tasks it takes over, but by the extent of agency, knowledge, and responsibility that a person develops through its use.
