Responsible AI Governance
Maturity Model & Strategy

Figuring out how to govern AI responsibly, maximize the benefits, and minimize the risks, is a barrier for many organizations.

For example, in a recent survey, 52% of executives said that they actively discourage generative AI adoption. The lack of a responsible AI strategy was the second most common reason.

These concerns are appropriate because the stakes for organizations are high, including risks and rewards to the quality of their products, reputation, client attraction, employee attraction, and compliance.

The stakes for end users, related communities, and society at large are also high, including mass disinformation, discrimination, privacy violations, and physical and psychological harm, to name just a few.

On this page, you will find a framework, case studies, and insights about evaluating and improving AI governance in companies that develop or deploy it.

The framework is based on Dotan et al. (2024), and builds on the NIST AI RMF, one of the top AI governance frameworks in the world.

All REPORTS

Flexible-Maturity-Model-for-AI-Governance - cover page1.png

Early-stage AI Governance

This IEEE report showcases the maturity model.

It contains the full questionnaire and scoring guidelines, as well as the rationale behine them.

Download

Early-stage AI Governance

In this hackathon, participants evaluated their early-stage projects, in the ideation stage

The evaluations included risk triage, governance assessment, and prioritizing next steps

Read examples of evaluations

Read insights from the hackathon, including top-ranked risks, common governance strengths and weaknesses, and common improvement priorities

Download

RAI Governance Hackathon Report cover.png

All Tech Is Human Hackathon

In this hackathon, participants used the maturity model to evaluate companies externally, based on publicly available documents.

Read examples of evaluations

Read insights from the hackathon, including scoring dilemmas, differences between evaluating companies of different sizes, and on what evaluations could be based.

Download

Case Study: Light-It

Light-It is a digital product agency building tailor-made healthcare web and mobile applications

They used the maturity model to evaluate themselves
Read reflections on the process

Adam Mallát, Innovation Manager: "[Filling out the questionnaire] was the first time, that I personally went into so much detail and became so analytical about what we do in the area of of ethics"

Download

The Full Academic Paper

Read detailed versions of the evaluation questionnaire and scoring guidelines

Understand the rationale behind them and how they are based on the NIST AI RMF

Learn about maturity models in AI ethics

Download

Who the model helps

Software Engineer

The guidelines can help companies that are implementing/utilizing/plan to utilize AI products and services to create their own AI Ethics blueprint

AI Ethicist

Auditors and regulatory bodies can utilize the questionnaire as a standardized tool for evaluating companies' AI governance practices. It provides a structured framework for assessing compliance with relevant regulations, standards, and ethical guidelines

Policy Professional

Small to medium-sized enterprises or startups that may not have the resources to appoint a dedicated person or team for AI governance can significantly benefit from these tools. The questionnaire and guidelines provide a structured approach to evaluating their current AI governance practices, helping these organizations identify areas of strength and weakness

How the model helps

Software Engineer

The experience was eye-opening for me. It was very helpful to have a framework to work off of. I often think about the topics that the framework walks you through, but until working with it I did not have a robust way of assessing the topics and a way to ground and level my assessment. I found it very helpful

Technology Consultant

The framework is a great sounding board for organisations at different phases of their AI journey. It enables them to catch issue[s] early in the system reducing cost and reputation implications

Tech Strategist

I think that the AI Governance Maturity Model raises many excellent questions that AI companies and projects should strive to address, and to do so in a requisitely comprehensive way. I value having the Maturity Model as a reference to return to as our project continues to develop, and as a tool to plan for future considerations