van der Schaar Lab

The van der Schaar Lab’s crucial role in AI safety and trustworthiness

The need for designing AI which is safe, trustworthy, and aligned with human objectives

In the light of the recent global AI Safety Summit, safety and trustworthiness are taking centre stage in public AI discourse. With the aim to consider the risks of AI and how to mitigate them, the summit brought together international governments, leading AI companies, civil society groups, and a large number of esteemed research experts. As pressing as the conversation seems to have become over the past weeks running up to the summit, the reality is that work on AI safety has been going on for a long time.

Professor Mihaela van der Schaar is one of the world-leading AI researchers that has been deeply involved in the UK’s search for a responsible yet transformative direction to take with AI, having previously advised the UK parliament on the governance of AI. In October, she was part of insightful panel discussions on AI safety at the Royal Society and Royal Statistical Society surrounding the AI Safety Summit.

With the rapid advance of AI and machine learning, closely monitoring and discussing opportunities and risks is essential. The van der Schaar Lab is driving innovation in the field but also conscious about the necessity of AI and ML tools being understandable and useable in a responsible, secure way. In essence, we are pursuing AI that works for humans:

In our relentless pursuit of a reality-centric AI that can solve complex real-world problems, we must consistently prioritise and uphold safety, trustworthiness, and alignment with human objectives. This critical trio serves as the foundational bedrock for AI technologies, ensuring they enhance human capabilities and contribute meaningfully to societal advancement.

Prof Mihaela van der Schaar

Our lab has positioned itself at the centre of the conversation through years of pushing the boundaries of safe and interpretable AI that works in the real world:

To learn more about our labs efforts, we encourage you to have a look at the pages above.

Andreas Bedorf