AI Assurance Masterclasses
Pioneering a holistic approach to certifying AI
Artificial Intelligence (AI) has seen remarkable advancements in recent years, with the rise of generative AI further increasing its accessibility and influence across various aspects of daily life. As this technology becomes more integrated into society, there is a growing need for greater awareness and education on effectively managing its risks.
AI assurance masterclasses provide a comprehensive understanding of the key principles involved in ensuring the safe and effective use of AI throughout its development lifecycle. They explore the evolving landscape of AI ethics and legal compliance both in the UK and globally, as well as the latest advancements in AI assurance.
These sessions focus on what aspects of AI need assurance and how to achieve it, highlighting the essential skills required within organisations and the key responsibilities of AI assurance teams. The ultimate goal is to promote the responsible deployment of both narrow and generative AI, ensuring alignment with organisational needs, ethical and legal standards, and technical best practices.
Our Partners
Validate AI, Imperial DSI, and the Operational Research Society have collaborated to deliver this cutting-edge masterclass. Drawing on our extensive experience working with organisations across academia, industry, and government, this program equips participants with the foundational skills needed to implement AI assurance efficiently within their businesses. Additionally, it encourages critical discussions on the future of AI assurance and its evolving landscape.
Some of the questions that we will address include:
- What evidence will key stakeholders require to enable trust in AI?
- What insights can we gain from previous experiences to design AI Assurance checks?
- What are the roles, responsibilities, and processes within organisations to deliver trusted AI?
- How can we engage with the wide array of stakeholders to build trust?
- How do we mitigate the risks of AI bias and poor prediction performance in sub-populations?
