COMPASS produces research-grounded analysis and recommendations on a wide array of issues: the legal status of AI; its social and environmental impacts; AI companies’ legal accountability for harms associated with their products; and national and international inequalities in AI development and deployment.

Rather than focusing on speculative future scenarios, COMPASS examines the concrete, real-world choices that need to be made today. It asks questions that receive insufficient attention: who benefits from different policy options? Whose voices are taken into account, both nationally and internationally? How should accountability be understood in the era of AI? What legal and policy options are needed to address the effects that AI is already having on individuals, societies, and the natural world?

As a commons, it brings together a global community of researchers and practitioners spanning law, the sciences, technology, the arts and the humanities, and civic leadership. Our outputs and funding are independent of industry and seek to advance the public interest.

The starting point of our work is that AI systems are human-made machines that must advance human and planetary well-being. That means holding companies accountable for the products they build, refusing to treat AI systems as legal persons, and requiring companies to design models that preserve a clear distinction between machines and people.

COMPASS focuses on technologies of automation (large language models, generative AI, algorithmic decision-making systems). Although our scholarly and practical expertise lies in the legal and policy domain, we collaborate with thinkers and doers from a wide range of fields to examine the broader societal consequences of these technologies, from their impact on human work and well-being to their effects on the natural world.