13 The GenAIUM framework

To guide ethical and effective AI integration, organisations can adopt the GenAIUM framework. This framework is designed to offer a quick start guide to considering the potential for using GenAI within organisations.

The GenAIUM Framework has three strands to it – assessment, interrogation and use.

Described image
Infographic created by K Barker using the AI prompt: Generate framework infographic - with the text on the infographic provided to AI.
Show description|Hide description

A visual framework titled GenAIUM FRAMEWORK is presented in bold, dark blue uppercase letters at the top against a light beige background. Below the title, three sequential stages are shown from left to right: Assess, Interrogate, and Use Model, each enclosed in a rounded rectangular box with arrows connecting them. Assess includes bullet points: Review the intended use case. Determine if the task is suitable for automation. Assess risks and define the level of required oversight. Interrogate includes bullet points: Examine the AI tool itself. Investigate training data sources, identify known biases, evaluate model transparency. Confirm legal compliance (e.g., UK GDPR, anti-discrimination laws). Use Model includes bullet points: Deploy AI responsibly, aligned with organisational values. Define fallback mechanisms for high-risk areas, such as reverting to manual processes or applying human arbitration in contested decisions. The layout is clean, minimalist, and professional, supporting clear comprehension of AI risk governance steps.

Infographic created by K Barker using the AI prompt: Generate framework infographic - with the text on the infographic provided to AI.

The Organisation for Economic Co-operation and Development (OECD, 2023) underscores that risk management must span the entire AI lifecycle to ensure trustworthy systems.

Session 3: Mitigating AI risks – 40 minutes

14 Developing a risk management plan