Erickson AI Assistant: Commitment to Confidentiality, Vision, and Purpose

At Erickson Coaching International, safeguarding the confidentiality and privacy of our learners and clients is not just a regulatory obligation—it's a deeply held value that guides our entire approach to integrating Artificial Intelligence.

We are building AI-powered tools to extend personalization, deepen engagement, and ensure coaching knowledge remains applicable, responsive, and truly human-centric. Our goal is to make powerful learning more accessible while remaining respectful of each individual's learning journey.

Marilyn Atkinson's Vision for AI and Coaching

Our founder, Marilyn Atkinson, has long championed innovation in coaching education. Her vision for AI is clear:

• AI as an Intelligence Builder: Marilyn sees AI as a tool to build intelligent systems that use analogy, metaphor, and meaning to support transformation. While AI cannot feel or embody human emotion, it can be guided to serve emotional understanding and resonance.
• The Erickson Advantage: We bring unique contributions to AI: the Coach Position, the Logical Levels framework, and solution-focused, generative questions. These elevate AI from basic assistance to transformational support.

• Strategic Partnerships: Erickson is engaging with partners like OpenAI and Anthropic to contribute high-quality, structured data that helps make AI more emotionally attuned and values-driven.

What We Use to Train the Erickson AI Assistant

To ensure high-quality, useful responses from the AI, we've trained the Erickson AI Assistant using over 600 hours of Erickson-produced and ethically sourced material. This includes:

• Faculty-led coaching demonstrations and videos shared on Canvas as core class content
• Mentoring and oral assessment sessions (only with documented consent from all parties involved)
• Internal documents like process guides, role descriptions, and coaching frameworks
• Business-building resources and Erickson+ community materials, excluding any participant who did not provide explicit consent

What We Do NOT Use

To protect privacy and meet the highest ethical standards, we do not use:

• Class recordings that involve learners or facilitators without explicit written permission
• Canvas content generated by students (e.g., comments, journals, assignments, audio uploads)
• Data from HubSpot, email, WhatsApp, Zoom, or any other platform involving student communication
• Any agreement data or personal records

Why We Chose a Custom GPT (Not a Branded Platform)

Rather than creating a proprietary Erickson-branded front-end, we intentionally chose to host the Erickson AI Assistant as a custom GPT within OpenAI's secure platform. Why?

• Data Protection: OpenAI adheres to world-class data security protocols, and Erickson is part of a Team Agreement with OpenAI that ensures all data remains private and is not used to train any public AI models.
• Simplicity and Impact: We focused on a 20% effort that delivers 80% of the benefit. Building a separate front-end with user authentication would have created access barriers without materially increasing security.
• Ethical Sharing: We believe in making solution-focused methodologies more widely available, and this platform enables safe, open access while respecting privacy.

Transparency Moving Forward

As we continue evolving our tools, any future applications (like AI-enhanced learning platforms) will be fully transparent. We will always disclose what data is used, how it's protected, and what choices learners have. Erickson is PIPA-compliant in British Columbia, Canada: https://www.bclaws.gov.bc.ca/civix/document/id/complete/statreg/00_03063_01 

Your Voice Matters

Have thoughts or feedback about the Erickson AI Assistant? We'd love to hear from you: https://www.surveymonkey.com/r/EricksonAIAssistant 

Try the Erickson AI Assistant

Access it here: https://chatgpt.com/g/g-682d134fa144819189ce15e745fffe49-erickson-ai-assistant 

We are committed to continuous improvement, responsible innovation, and learner-centered design.