Climate justice and trustworthy AI: policy foresight workshop

"Soap bubble", mezzotint by M. Rapine, after an image by Alexandre-Blaise Desgoffe; plate VI in Le monde physique (1882)

This workshop is an initiative from Mozilla’s Environmental Champions in Mozilla’s Sustainability Program and will be facilitated by Michelle Thorne and Fieke Jansen.

The world will cross the 1.5 degree warming threshold in 2024, quicker than previously estimated, forecasts the UN World Meteorological Organization. We have three and a half years to dramatically cut our emissions. We need sustainable systems, and we need them now. 

We describe sustainability as a healthy environment, economic well-being, and social connection. We are particularly interested in reducing the internet’s significant emissions while advocating to keep this global public resource open and accessible to all. It is essential that the internet advances healthy, sustainable practices. 

Increasingly, our online lives are affected by artificial intelligence systems. If we want a healthy internet—and a healthy digital society—we need to make sure AI is trustworthy. For AI to be trustworthy, we need AI that is demonstrably worthy of trust. Privacy, transparency, and human well-being are key considerations, and there is accountability for harms.

We must meet the moment of the climate crisis now—in the era of AI. This means bringing together two currently separate conversations: mitigating climate change and addressing the civil rights concerns in AI systems. We must shift from a discourse on greenhouse gases and melting ice caps to a civil rights movement with the people and communities most vulnerable to climate impact and technology at its heart. The movement for climate justice and the call for trustworthy AI need to be linked. Independently, they have made great gains by positioning harms not as solely environmental or technical shortcomings but as justice issues: bans on facial recognition, ad tech reform, landmark surveillance cases and youth climate lawsuits to name a few. 

To link trustworthy AI to climate justice, we must expand our understanding of human well-being and AI harms. Research demonstrates how AI intensifies energy consumption and AI systems developed by major tech companies (Amazon, Microsoft and more) are used to speed up extraction of oil and other natural resources. Tech companies are announcing ambitious climate plans, often following pressure and mobilization from their workforce. But none of these efforts take full account of the harms caused by their AI systems

Our proposal

To understand the environmental harms caused by AI and situate these in a larger context of internet health and the climate crisis, we are convening an interdisciplinary group. We seek to demand AI systems that are trustworthy and sustainable. “Collective liberation and ecological sustainability,” as described by Sasha Costanza-Chock in their book Design Justice, will be a guiding vision.

With these findings, we will develop a stronger call for sustainable and trustworthy AI that translates into policies and organizational agendas. We will pursue constructive approaches along the way, such as sustainability engineering and sustainability by design, aiming to create blueprints for others to build on.

We will focus this first convening on the policy window in Europe as the European Commission defines its AI strategy, climate agenda and COVID-19 recovery plans. This moment offers a strategic opportunity to advance trustworthy AI and climate justice in the summer of 2020. 

If you would like to share your perspective on these issues or host similar conversations, please email me at michelle at mozillafoundation dot org or Mozilla’s sustainability program at

Image: “Soap bubble”, mezzotint by M. Rapine, after an image by Alexandre-Blaise Desgoffe; plate VI in Le monde physique (1882). Source: Public Domain Review.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.