What the IOC's Trustworthy AI Framework asks sports organisations to do

The International Olympic Committee (IOC) said on 28 September 2026 that it "has launched the first Trustworthy AI Framework for the Olympic Movement", a 22-page guidance document with an accompanying self-assessment checklist. We read the IOC's announcement and the full framework document; this is what it asks sports organisations to do.

Who it is for

The IOC says the framework "has been created for organisations and stakeholders across the Olympic Movement, from those beginning to explore AI to those already integrating it into their activities." The document itself says it was created "to help athletes, officials and organisations across the Olympic Movement explore new AI technologies responsibly and with confidence."

The seven principles

The framework is "built upon seven Trustworthy AI Principles", listed in this order:

1. Prioritise safe and secure AI for sport
2. Protect personal data with respect and care
3. Deliver robust and reliable AI for sport
4. Make sure the use of AI is clear and transparent
5. Govern AI with accountability and responsibility
6. Ensure AI is fair and equitable for all
7. Support sustainable AI for the future of sport

Each principle comes with "key ambitions". Examples from the document include to "Collect only the minimum necessary personal data"; to "Tell athletes, officials, staff, volunteers and fans when AI systems have access to their personal data and for what purpose"; to "Ensure that qualified individuals review or supervise AI outputs if their decisions could affect competition integrity, athlete welfare and/or organisational operations"; to "Develop AI incident response and action plans"; and to "Regularly test for bias in datasets and AI models".

The IOC's announcement says the principles "can be applied throughout the lifecycle of an AI system, from initial design and procurement through to implementation and ongoing use."

The checklist

The document describes the Trustworthy AI Checklist as "a three-step process":

• Step 1, evaluate risk. An "Inherent Risk Assessment" of "high-level, multiple-choice questions" leads the tool "to suggest an inherent risk rating of low, moderate or high". The document adds that a "high" rating "should not necessarily rule out the use of the AI system or application in question", but "signals the need for rigorous controls, deep scrutiny and – potentially – a more extensive mitigation strategy".
• Step 2, embed trust. Users answer "around 70 multiple-choice questions in seven sections inspired by each of the seven principles", and the tool suggests "a series of automatically generated ratings".
• Step 3, enhance with recommendations. The tool "delivers a series of automatically generated recommendations" in four parts: a top-line recommendation per section, action points, links to resources, and suggestions for collaborators. (The IOC's announcement labels this step "Improve their approach".)

The checklist is published as a separate spreadsheet alongside the document and a training pack; we have not reviewed either.

Binding or not?

The documents do not use the word "voluntary". What they do say:

• The IOC's announcement: "Responsibility for the trusted ethical development and use of AI remains with the organisations that design and/or deploy these systems".
• The document describes itself as "an invitation to embed the Trustworthy AI Principles and Checklist into organisations' AI strategies", and says the checklist "is designed to support self-assessment and internal monitoring."
• Its disclaimer says the framework "is intended to serve as practical guidance" and "is not, and should not be construed as, a legal document or a substitute for legal advice." The same disclaimer also states: "All organisations are expected to design and use AI systems responsibly and to the highest standards".

The document says the framework "will undergo periodic reviews and updates".

How it was developed

The IOC says the framework "was developed by the IOC with the support of Digital Transformation Consulting Partner Deloitte and shaped through engagement with more than 200 leaders from across the Olympic Movement. This included input from athletes and experts across technology, governance and sport." The document refers to Deloitte as a "Worldwide Olympic Partner" and says the 200-plus leaders came "from International Federations (IFs), National Olympic Committees (NOCs), Games hosts, athletes and others".

What officials said

Alejandro Merino-Madrid, the IOC's Head of Artificial Intelligence, said: "But technology must always serve sport — not the other way around. That means protecting athlete data, safeguarding fair competition and ensuring people understand how AI is being used."

Olympic medallist and IOC Member Sarah Walker said: "Athletes helped build this Framework, and it helps people make good choices about AI, ones that are open about how decisions get made and keep the athlete at the centre."

Mark Price of Deloitte said the firm had "helped develop practical guidance that enables organisations across the Olympic Movement to capture AI opportunities, while navigating risks to make informed decisions".

Background

The framework document says the Olympic AI Agenda was "published by the IOC in April 2024" and "was created to guide the Olympic Movement in its use of new AI technology." The IOC's Olympic AI Agenda page says that agenda "sets out a governance and oversight framework to identify and mitigate risk".

Why it matters

On our reading, the framework is guidance rather than regulation: the IOC places responsibility with the organisations using AI and offers a self-assessment tool, not an approval process. The document says "many sports organisations lack the know-how, the resources, the capabilities or the infrastructure" to develop or implement AI; for them, a shared checklist may be the more practical part of the package.