A workshop on AI, societal stability and long-term value for investors and others attending PRI in Person 2026

Investors are now heavily exposed to artificial intelligence: directly, through the companies building the models, chips, data centres and energy supply behind it, and indirectly, through every company and economy the technology reaches. The long-term value of those holdings depends on how society responds to AI, through regulation, public consent, labour markets and the resilience of its
institutions, and on what markets expect that response to be. Investors need a way to assess that dependence and to act on it.

Silicon Boundaries, a joint initiative of the Oxford Sustainable Finance Group and Theia Finance Labs, offers one. Drawing on the idea of planetary boundaries, it sets out to identify the thresholds beyond which the benefits of AI and related technologies are outweighed by their contribution to societal instability, across nine categories that include information integrity, social cohesion, financial stability and
the environment. Four scenarios (Goldilocks, Luddite, Gilded Cage and Icarus) describe how the technology and society’s response to it might develop, and what each path would mean for the companies and portfolios investors hold.

After a short introduction to the framework, we will work with participants on the data needed to apply it, looking at what companies disclose today, which indicators could track each boundary, and where the gaps lie. We will then turn to the products that data would support, such as portfolio exposure analysis, company assessments and stewardship toolkits, and identify where to start. After the break, we will explore pathways to implementation for investors, from investment beliefs and strategic asset allocation to manager selection, mandates, stewardship and policy engagement. We want to draw on participants’ experience and hear what investors would need to
act on these risks, and which tools they would use.

To apply for a place, please reply to Philip Stevens by Monday 12 October, including your name, position, and affiliation. Places are limited and will be allocated to asset owners first. Applications will be considered case by case.