Module 22- Video w/ Rebecca Rutschmann- the role of AI in health coaching

Definition and Distinctions in AI and Coaching

AI in coaching and AI coaching are two distinct concepts that define how artificial intelligence interacts with the coaching process:

AI in Coaching: This refers to AI augmenting a human-led coaching process. Artificial intelligence serves as a component within the workflow, but the overall process remains under the direction of a human coach. Examples include using AI note-takers, utilizing AI for analysis, or using it to prepare for sessions.
AI Coaching: This is a fully AI-driven, autonomous system. It operates without human involvement and interacts with the client or consumer on its own terms.

The Three Pillars of AI Systems

AI systems are constructed upon three primary pillars, often compared to the process of cooking:

The Recipes (Algorithms): These are the mathematical formulas and instructions. They have been in existence and under development since the $1960s$ .
The Ingredients (Big Data): High-quality outcomes require high-quality data. The "big data boom" occurred in the February (following the development of the internet in $1997$ ), allowing companies to collect and make sense of massive amounts of information to fuel algorithms.
The Kitchen Equipment (Computing Power): Powerful hardware is necessary to process large datasets and execute complex algorithms. It is only since the February that computers have become powerful enough to generate the advanced algorithms and generative AI available today.

Generative AI Mechanics and Hallucinations

Generative AI operates primarily through next-word prediction. It assesses the probability of the next token in a sentence (e.g., predicting if the sky is "blue" or "dark").

Hallucinations: This occurs when AI outputs information that is untrue. Because AI is designed to predict the next probable word, it will generate an output even when it lacks clear references.
Research Example: A researcher inquired with Chad Chibiti about other researchers in his field. Chad Chibiti provided a specific list of names, but the researcher discovered they did not exist; the AI had fabricated them because no other AI coaching researchers existed in its database.
Implications: Hallucinations are considered a feature of the software's predictive nature, not a bug. However, in health coaching, these fabrications can be dangerous if a client receives false medical or wellness information.

AI in Health and Wellness: Current Trends and Risks

Approximately $230,000,000$ people per week discuss their health with Chateapiti. Activities include uploading medical records, test results, and connecting wellness applications.

Specific Tools: Spruce is the Chattypiti voice feature on Open the Eyes Chattypiti. In one instance, a user with extreme abdominal pain used Spruce to stay calm while waiting for a doctor's appointment. The user, a single mom with one child, later used CHEAPT as a partner to understand her surgery (appendix removal) and establish healthy habits post-operation.
Regulatory Status: ChatGPT Health, launched in January $2026$ , is not a healthcare provider. It is not covered by federal health privacy laws such as HIPAA in the United States or GDPR in Europe.
Professional Versions: CHPBT in card for health care operates in a protected environment with safeguards for HIPAA compliance, but this is not available to the general consumer.
The Second Voice: Coaches are increasingly becoming the "second voice in the room," helping clients navigate the boundary between general information and misleading medical advice produced by AI.

The Four Pillars of Responsible AI

When these four pillars are maintained, the result is trust between the system and the user:

Ethics: The anchor of the system, focusing on fairness and accountability.
Explainability: Ensuring the user understands the "why," "what," and "how" of the AI's operations.
Privacy: Respecting consent and handling data with care through frameworks like GDPR and HIPAA.
Security: Protecting AI systems against misuse or data breaches.

Governance and Ethical Guardrails in Coaching

Responsibility for AI is shared across different entities:

Governing Bodies: Organizations like the National Board of Health and Wellness Coaching are responsible for codifying ethics.
Developers: They must design for security and explainability.
Coaches: Individual practitioners are responsible for safeguarding client privacy and disclosing AI use. The code of ethics requires disclosing the use of artificial intelligence and emerging technologies, including how they affect confidentiality.

Global Consent Laws for Recording and Data Use

Consent requirements vary significantly by region:

One-Party Consent: Only one participant in a conversation needs to agree to the recording.
All-Party Consent: All involved parties must agree to the use of recording tools.
European Union: Requires strictly proven, freely given, specific, and unambiguous consent.
United States and Australia: Consent laws depend on the specific state of residence (one-party vs. all-party).
Asia and Middle East (United Arab Emirates): Recording without explicit knowledge is strictly forbidden.

Integration Fields for AI in Coaching

Effective use of AI in coaching covers three main areas:

1. AI as a Business Ally

Thinking Partners: Tools like ChatGDP, Cloud, Gemini by Google, Mistral (useful for European languages), and Microsoft Copilot can be used for brainstorming.
Research and Learning: Perplexity and Notebook LM aid in self-learning.
Content Creation: AI helps generate text for emails, website copy, course outlines, images, and videos (using tools like Loom, Canva, or Zoom).
Avatars: AI avatar presenters can be used to explain complex content to clients.

2. Enhancing the Client Experience

Session Preparation: Creating AI assistants for onboarding, contracting, and defining coaching goals.
During Sessions: Role-playing games for behavior change, brainstorming, and change talks.
Between Sessions: Providing accountability, journaling prompts, and self-reflection. Clients often find AI chatbots non-judgmental, making it easier to discuss sensitive topics.

3. AI for Personal Growth

Reflective Practice: Using compliant systems to process anonymized transcripts to identify energy shifts or improve active listening.
Supervision: Using AI as a supervisor for self-reflection.
Communication Skills: Simulating challenging conversations to receive feedback on filler words, pauses, eye gaze, and posture.

Four-Point Compliance Checklist

To ensure safe AI usage, practitioners should verify the following:

Data Storage: Identify where data is processed and ensure clients can request data deletion.
Mutual Consent: Confirm consent for recording at the start of every session and determine what happens to data if consent is withheld.
Training Models: Check if the data entered is used to train AI models; ensure client data is excluded from model training, especially in free versions of tools.
Transparency: Mention all AI tools explicitly in the coaching contract.

Three-Step Action Plan for Coaches

Step 1: Establish ethical guardrails and a safety baseline to protect session integrity.
Step 2: Use AI for personal productivity first to understand how the systems function.
Step 3: Make informed decisions on business tools and update the consent framework in coaching contracts.

Digital Etiquette for AI Coaching

Contracting: Partner with the client to decide which AI tools they are comfortable with. Use tick-boxes in contracts to allow clients to consent to specific tools while opting out of others.
Anonymization: Clean all files of names, birth dates, and personal information before uploading them to any AI tool.
Recontracting: Verify at the start of every session if the client still feels comfortable using AI, as their preference may change depending on the vulnerability of the day's topic.

Note on Recorded Remarks

The following verbal expressions were noted at the conclusion of the material: "Oh my god. This shit. Even if you don't wanna go, just say something like oh my god. You're like oh shit."