Magif is designed to extend coaching responsibly: protecting your methodology, supporting clients between sessions, and keeping professional judgment with the coach.
In 2025 the International Coaching Federation (ICF) published its Artificial Intelligence (AI) Coaching Framework and Standards — a guide for using AI in coaching ethically and effectively. Rather than starting from scratch, the framework builds on the ICF Core Competencies, the same foundation used to assess human coaches, and adds guidelines specific to AI and software.
The standards map core AI coaching capabilities such as active listening and goal tracking to specific requirements, organized by domain and by complexity (basic and advanced). They come with complementary resources: worked Examples, and a set of “Questions Stakeholders Should Ask” before trusting an AI coaching system.
Coaching depends on trust, confidentiality, and methodology. AI should support that work responsibly, not blur those boundaries.
Your data is never used to train anyone else’s AI
Each AI stays isolated to your content and instructions
Client conversations stay private
Data is stored in the EU
We do not position AI as a replacement for the coach. We design it as a responsible extension of your methodology, client support, and offer delivery.
The trust layer behind Magif comes from research, standards work, and long-term involvement in AI coaching.
Built with research leadership in AI coaching and leader development.
Research summary on HAL →Magif.ai's cofounder co-authored ICF AI coaching standards.
AI Coaching: Standards for Ethical and Accessible Coaching →Built within the AI coaching and professional standards ecosystem, not outside it.
Meet the founders →Our ongoing contributions to the AI coaching ecosystem through academic research, international conferences, and professional standards.
The questions coaches actually ask before trusting AI in their practice.
Still have questions?Contact our support team
Use this page to understand how Magif thinks about standards, privacy, and responsible AI use before launching a client-facing AI agent.