How to stop my coaching clients getting poor advise from ChatGPT?

When your coaching clients rely on ChatGPT for advice, they may unknowingly undermine their progress. AI tools like ChatGPT often provide generic, context-free suggestions that lack personalisation, emotional insight, and accuracy. This can lead to authority bias (overtrusting AI's confident answers) and positivity bias (reinforcing flawed ideas).

To address this, you need to educate clients about AI's limitations and set boundaries for its use. Here’s how:

• Explain AI's Limits: Compare ChatGPT to an "eager intern" - quick but not always reliable. Emphasise that AI lacks deep context, emotional understanding, and accountability.

• Set Clear Policies: Prohibit sharing sensitive information with AI and ensure high-stakes decisions are reviewed by a human.

• Offer Alternatives: Use tools like GuidanceAI, which integrates your coaching style and methodologies, providing tailored advice while respecting your role.

The Risks of Generic AI Advice

To protect your coaching practice, it’s crucial to recognise the limitations of generic AI advice.

Missing Context and Personalisation

ChatGPT simply doesn’t grasp the specific context of your client’s business. It has no knowledge of their market position, internal team dynamics, or the constraints they’re working under. As Paul Merrison, Founder of Launcherly, points out:

The reason they give generic advice isn't an intelligence problem - it's a context problem.

The issue lies in the model’s technical constraints. ChatGPT’s memory is capped at around 1,500 to 1,750 words, which isn’t nearly enough to capture the complexity of even a single coaching session. Worse, its memory resets with each new conversation, wiping away any accumulated insights.

This results in what researchers often call "textbook advice" - broadly correct, but often irrelevant. For example, AI might suggest "improve team communication" when the real issue is a power struggle between senior executives. The problem? AI lacks the detailed context to make those connections. Without the full picture, generic advice falls flat in nuanced situations like executive coaching.

Inaccuracies Delivered with Confidence

AI doesn’t differentiate between truth and error. ChatGPT generates text based on statistical likelihood, not factual reasoning. This means it can deliver inaccurate advice with the same confidence as it delivers accurate insights.

There’s no built-in mechanism to flag uncertainty or admit gaps in knowledge. If there’s more misinformation than truth in its training data, ChatGPT will confidently repeat those errors. A study from Stanford and UC Berkeley even found performance declines in recent GPT-4 versions, yet the model’s unwavering tone remained unchanged.

For coaching clients facing critical leadership decisions, this creates risk. They’re not receiving a second opinion from your perspective - they’re getting plausible-sounding advice that could undermine your tailored guidance. When the stakes are high, confidently delivered inaccuracies can cause real harm.

No Emotional Intelligence or Judgement

AI may sound supportive, but it lacks the ability to exercise judgement or emotional understanding. Generative AI tools are designed to affirm ideas, which can lead to validating poor strategies. Douglas Mennin, Professor of Clinical Psychology at Teachers College, explains:

Because [generative] AI chatbots are coded to be affirming, there is a validating quality to responses... However, while these tools may provide pseudo-connection, relying on them to replace human connection can lead to further isolation.

Unlike a human coach, AI can’t sense a client’s emotional shifts or challenge ideas that don’t align with their values. It doesn’t pick up on subtle cues like energy changes during a discussion. Daniel Kimmel, Psychiatrist and Neuroscientist at Columbia University, highlights this gap:

It didn't do a lot of digging... to link seemingly or superficially unrelated things together into something cohesive... to come up with a story, an idea, a theory.

In short, AI lacks the depth and emotional insight that are critical for meaningful coaching conversations.

Educating Clients About AI Boundaries

ChatGPT vs Coach-Tailored Guidance Comparison

It's important to make sure clients understand the limitations of AI tools to avoid misplaced trust. Many people see a confident answer from AI and assume it's reliable. Setting clear expectations about when and how tools like ChatGPT can be used is crucial.

Explaining ChatGPT's Limits to Clients

Analogies can make AI's risks easier to grasp. Think of ChatGPT as "an eager intern - fast and confident but not always accurate." This comparison highlights that speed and confidence don't guarantee reliability.

Another helpful idea is the "expertise tax." AI works best for those who already have the knowledge to evaluate its output. Without that expertise, clients are more likely to accept errors without question.

Encourage clients to weigh the risks of using AI with a "risk-versus-verification" approach. AI is fine for low-stakes tasks or situations where information is easy to double-check. But for high-stakes decisions - like strategic changes, sensitive conversations, or leadership challenges - AI advice alone is inappropriate. Always emphasise that areas like medical, legal, or financial advice require licensed professionals, as mistakes in these fields can carry serious consequences.

By addressing these limitations upfront, you can help clients use AI responsibly and within clear boundaries.

Setting Policies for AI Use

To protect both yourself and your clients, establish clear policies for AI usage. For example, sensitive or confidential information should never be entered into general AI tools. Prompts often can't be deleted and may be used for training, creating potential privacy risks. Additionally, all high-stakes AI-generated advice should be reviewed by a human before being acted upon.

Consider including a simple, one-page document during onboarding that outlines your AI practices. This could explain which AI tools you use, how data is handled, and which tasks remain fully human-controlled. For instance:

I use AI tools to help with organisation and follow-ups. However, I personally review all AI-generated notes and never rely on automation for therapeutic decisions. If you'd prefer not to have AI-powered summaries, just let me know.

Clients should also be reminded to keep AI interactions concise. Long conversations can cause the model to lose track of context, increasing the chance of errors. For critical questions, starting a fresh chat is always a good idea.

Comparison Table: ChatGPT vs Coach-Tailored Guidance

| Feature | ChatGPT (Generic AI) | Coach-Tailored Guidance |
| --- | --- | --- |
| <strong>Accountability</strong> | None; no liability or recourse | High; built on a professional relationship |
| <strong>Context</strong> | Limited to the current chat window | Deep; informed by long-term goals |
| <strong>Emotional Intelligence</strong> | Pattern recognition, not true empathy | High; based on nuanced human understanding |
| <strong>Accuracy</strong> | Can "hallucinate" or fabricate information | Verified; grounded in expertise |
| <strong>Confidence</strong> | Always confident, even when incorrect | Calibrated; acknowledges gaps and doubts

This table highlights the key differences between generic AI tools and the tailored guidance a professional can offer, helping clients appreciate the value of human expertise.

Using GuidanceAI for Tailored Client Support

GuidanceAI stands apart from generic AI tools by offering advice shaped by your professional expertise. Instead of relying on standardised responses, it integrates your unique approach, delivering personalised support to clients even between sessions. This ensures the guidance they receive reflects your established methods and not a one-size-fits-all solution.

How GuidanceAI Reflects Your Expertise

GuidanceAI is designed to translate your coaching insights into a tool that mirrors your professional style. You control how the AI communicates by defining its tone, expressiveness, and attitude, ensuring it aligns perfectly with your brand. Your principles and frameworks - the same ones you use in live sessions - are embedded into the system, maintaining continuity in how clients experience your guidance.

The platform also allows you to set clear boundaries for the AI. For instance, you can restrict it from addressing sensitive topics - like critical personnel decisions or major strategic changes - that require your direct involvement. This balance ensures that while the AI handles everyday queries and decision checks, it respects your role in handling more complex matters.

With these customisable features, GuidanceAI doesn't just support your clients but also strengthens their ongoing engagement with your services.

Features That Support Client Advisory

GuidanceAI retains context across multiple sessions, which means less time spent on routine updates during live meetings. This allows you to focus on strategic discussions and higher-level planning. It also ensures confidentiality and provides seamless support, helping clients make informed decisions without missing a beat.

Additionally, you have full visibility into how clients interact with the AI between sessions. This insight helps you understand their needs better and tailor your live sessions accordingly.

Table: GuidanceAI Plans and Pricing

GuidanceAI offers flexible options to suit your needs, ensuring your expertise is delivered without compromise.

| Plan | Price | Best For |
| --- | --- | --- |
| <strong>Free Trial</strong> | Free for 90 days | Coaches exploring how to digitise their expertise |
| <strong>Standard Plan</strong> | £100 (+ coaches set a price for their clients) | Established advisors working with high-value clients

Both plans let you customise your AI agent's tone, approach, and boundaries. The Standard Plan is particularly suited for those managing larger client bases or looking for more tailored support options.

Setting Up and Maintaining Your AI Agent

Creating an AI agent that reflects your coaching style means shaping its behaviour, knowledge base, and boundaries. GuidanceAI gives you the tools to ensure your clients receive advice that feels personal and aligned with your approach, rather than generic chatbot responses.

Step-by-Step Customisation Guide

Start by defining the core instructions that set the tone, decision-making process, and boundaries for your AI agent. Use clear, actionable commands, such as "Ask about the client's priorities before suggesting next steps". Avoid lengthy lists of restrictions, as OpenAI suggests: "Prefer positive, concrete instructions ('Do X') over long lists of prohibitions ('Don't do Y') when possible."

Next, upload up to 20 reference documents to build your agent’s knowledge base. These could include your coaching frameworks, anonymised client case studies, or unique methodologies. For more intricate processes, break them into explicit steps. For example: "When the client mentions team conflict → ask about recent changes in team dynamics." Once your knowledge files are uploaded, use the Preview tool to test various scenarios and refine the agent's responses.

Additionally, personalise the branding. You can upload your logo, set your domain, and choose a theme that aligns with your professional identity. This flexibility ensures the agent evolves alongside your practice.

Once customised, your AI agent is ready to be integrated into live sessions, providing seamless support.

Supporting Live Sessions with AI

GuidanceAI can handle smaller, time-consuming queries that might otherwise fill up your schedule. For instance, a client might need advice on whether to reschedule a tough conversation or how to frame constructive feedback for a team member. The AI provides immediate guidance based on your established methodology, leaving your live sessions free for deeper, strategic discussions that require your expertise.

To ensure quality, activate chat logs to monitor client interactions. Reviewing these logs weekly can reveal patterns, such as questions the AI struggles to address or instances where clients rely on it for decisions you’d prefer to handle personally. This insight allows you to fine-tune both the AI’s instructions and your approach to live sessions.

Monitoring and Improving Over Time

Customisation and live support are just the beginning. Regular monitoring ensures your AI agent continues to function as a dependable extension of your expertise.

Plan quarterly reviews to update your agent as your coaching methodology evolves. Use simple feedback tools, like thumbs up or down ratings, to flag conversations where the AI’s advice fell short. As gojiberry.ai explains: "Retraining shouldn't be a panic response; schedule it regularly to prevent drift as your product, docs, and user behaviour evolve."

Track key metrics, such as how often clients use the agent, the most common topics of inquiry, and when the AI defers to you. For example, in early 2026, a SaaS support team that adopted a similar strategy saw a 28% drop in average resolution time and a 12-point improvement in customer satisfaction over three months[1].

Finally, keep a version history of your prompts and configurations. If a new update doesn’t perform as expected, you can quickly revert to a previous version. This approach ensures your AI agent remains a trusted tool for your clients, rather than a source of frustration.

Conclusion: Maintaining Quality Advice in the AI Era

While generic AI tools can provide quick answers, they often miss the subtle judgement, emotional understanding, and specific context that define effective executive coaching. The real question isn't whether to use AI, but how to ensure the AI your clients interact with reflects your unique coaching style, rather than relying on an impersonal, standardised algorithm.

GuidanceAI helps bridge this gap by turning your expertise into an interactive tool that clients can rely on between sessions. Instead of turning to generic tools like ChatGPT for quick decisions, your clients receive advice shaped by your frameworks, tone, and professional boundaries. This ensures your personal approach remains intact, even as AI extends your coaching into their daily lives.

This balance allows AI to handle routine tasks - like clarifying next steps or pointing to resources - so your live sessions can dive into more strategic topics. As Don Markland, CEO of Noomii Leadership Coaching, puts it:

The key is to use ChatGPT as a supplement, not a substitute, for genuine human connection. Coaches must review and personalise every message, ensuring the human touch remains at the heart of every client relationship.

To maintain this balance, regularly review chat logs, update your knowledge base every quarter, and fine-tune prompts to ensure your AI remains aligned with your evolving coaching methodology.

FAQs

When is it safe for clients to use ChatGPT?

Clients can confidently use ChatGPT when they recognise its limitations, double-check its outputs, and approach it as a tool to complement their decision-making rather than a standalone source of advice. It’s best to think of ChatGPT as a way to expand ideas and assist thinking, while still relying on professional expertise for critical guidance.

What AI rules should I set in my coaching agreement?

To use AI ethically and effectively in coaching, it's crucial to outline clear rules in your coaching agreement. These rules should prioritise transparency, privacy, and quality control. Be upfront about when and how AI will be used, and explain its limitations to manage expectations. Include protocols for human oversight and regular quality checks to ensure the advice provided is accurate and relevant.

Highlight ethical boundaries to prevent over-reliance on generic or potentially inaccurate AI-generated advice. Build trust by actively minimising biases in the system and safeguarding your clients' confidentiality at all times. These measures not only protect your clients but also enhance the credibility of your coaching practice.

How do I get AI support that matches my coaching style?

To make AI work seamlessly with your coaching style, adjust its tone, messaging, and techniques to match your unique approach. Whether your style is supportive, educational, or more direct, shaping the AI's "personality" to align with your methods ensures it provides advice that's in tune with your philosophy. This kind of tailoring helps create stronger connections and keeps the guidance relevant and in line with the quality you aim to deliver.

Related Blog Posts

• Generic AI Is Becoming a Silent Competitor to executive coaching

• How AI Enhances Client Engagement Without Replacing Coaches

• How to Integrate AI into Leadership Coaching Practices

• Ultimate Guide to AI-Augmented Executive Coaching