Technical AI Governance Workshop

As the development and use of AI systems expands, policymakers increasingly recognise the need for targeted actions that promote beneficial outcomes while mitigating potential harms. Yet there is often a gap between these policy goals and the technical knowledge required for effective implementation, risking ineffective or actively harmful results. 

To address this issue, we are hosting a workshop on Technical AI governance—a nascent field focused on providing analyses and tools to guide policy decisions and enhance policy implementation. With this workshop we aim to provide a venue that fostering interdisciplinary dialogue between machine learning researchers and policy experts. 

Technical AI Governance is a broad field encompassing a range of distinct subareas. Reuel et al. taxonomise the field according to a set of ‘capacities’, which can apply across the AI value chain. These are:

• Assessment: The ability to evaluate AI systems, involving both technical analyses and consideration of broader societal impacts;

• Access: The ability to interact with AI systems, including model internals, as well as obtain relevant data and information while avoiding unacceptable privacy costs;

• Verification: The ability of developers or third parties to verify claims made about AI systems’ development, behaviours, capabilities, or safety;

• Security: The development and implementation of measures to protect AI system components from unauthorised access, use, or tampering;

• Operationalisation: The translation of ethical principles, legal requirements, and governance objectives into concrete technical strategies, procedures, or standards;

• Ecosystem Monitoring: Understanding and studying the evolving landscape of AI development and application, and associated impacts.