Green AI Summit: Rethinking AI for a Sustainable Future

A collection of visionaries, scholars, and innovators came together at Harvard University and Boston University on April 25 and 26 for the Green AI Summit, focusing on the evolving relationship between artificial intelligence and environmental responsibility. The summit was organized by the Green AI Institute and Harvard Undergraduate AI and Sustainability Group, with support from the Boston University Center for Information and Systems Engineering (CISE).

As artificial intelligence becomes more deeply embedded in everyday life, questions about its ecological footprint have become more prominent. 

The summit addressed this growing concern by examining how AI can evolve to align with environmental priorities. Sessions and speakers covered various topics, from energy-efficient computing to policy frameworks that support sustainable AI development. 

A diverse slate of speakers—from tech executives to university researchers and public policy leaders—led a packed schedule of keynote talks, panels, and workshops. 

Among the major themes were balancing innovation with ecological restraint, leveraging AI to combat climate change, and ensuring equitable access to green technology.

Major themes included balancing technological innovation with ecological restraint, leveraging AI to combat climate change, and ensuring equitable access to green technology. Keynote speaker Jennifer Turliuk, Practice Leader in the Climate and Energy AI group at the Martin Trust Center for MIT Entrepreneurship, highlighted the need to prioritize AI projects where “the climate benefits exceed the climate harms” and advocated for frameworks that align AI’s capabilities with positive business and climate outcomes.

Throughout the summit, attendees discussed various topics, including carbon-aware computing, the ethical deployment of AI tools, and integrating machine learning into sustainable infrastructure design. 

Day One opened with remarks from Massachusetts Senator Ed Markey:

“The climate crisis is a five-alarm fire. 2024 was our hottest year ever on record,” he said. “As we face attacks on climate programs, it is more critical than ever that we fight for a just and livable future for everyone.” 

Markey added that artificial intelligence can be a tool to solve that problem.

“Training and deploying AI models consumes substantial power, massive amounts of electricity… To meet this moment, we need responsible long-term planning to address AI’s impact on our environment,” he said.

After the opening remarks, four panel discussions followed. 

Panel 1: AI and Data Center Energy Usage and Environmental Impact kicked things off by sharing that AI is power-hungry. Speakers explained how data centers—essential to running large-scale AI models—burn through electricity and water. The conversation centered on practical ways to cut emissions, from tapping renewable energy to rethinking cooling systems. The takeaway? If AI is going to scale responsibly, its infrastructure needs a serious sustainability upgrade.

The focus shifted to innovation in Panel 2: Towards a Low-Carbon AI and Data Center Ecosystem. Panelists explored what a “low-carbon” future might look like for AI, spotlighting energy-efficient design, more innovative tech infrastructure, and the policy support needed to make it happen. The session doubled down on the urgency of innovation—not just for efficiency but to prevent AI’s carbon footprint from losing control. 

Panel 3: Next-Generation Computing Hardware: Electronic, Optical, and Quantum Technologies took things further into the future. The panel dove into bleeding-edge hardware and asked how these technologies might go mainstream. Speakers addressed big-picture questions about performance, scalability, and environmental trade-offs while hinting at the tech that could redefine sustainable AI.

Finally, Panel 4: Green AI and Data Center Policy and Investment brought it back to the real world. This panel was all about who pays, who regulates, and how the pieces fit together. Speakers laid out the policy and investment strategies that could make Green AI more than just a buzzword, from public-private partnerships to government incentives.

Across the board, the panels revealed a complex—but promising—road ahead. 

Building a greener AI future isn’t just about better tech; it’s about collaboration across sectors, bold policy shifts, and reimagining what sustainable innovation looks like from the ground up.

Day 2 of the Summit shifted the spotlight to Boston University, where three insightful panels built on the momentum from Harvard’s previous kickoff event. 

With a full crowd at the Center for Computing and Data Science, the Saturday sessions leaned into applied solutions, real-world integration, and the broader global implications of sustainable AI development.

Panel 5: Efficient and Responsible AI: Technology Solutions for a Greener Future opened the day’s discussions by focusing on machine learning itself. 

Panelists spotlighted the backend of AI, including training models, storing data, and running algorithms. They shared their thoughts on how rethinking these processes could dramatically cut energy use. From optimizing neural networks to improving hardware-software synergy, the conversation centered on how efficiency doesn’t just save resources, it’s necessary. 

Panel 6: AI and the Grid – Complex Interactions followed, which dug into the relationship between AI systems and modern energy grids. Panelists explored how AI can help balance supply and demand, integrate renewable energy more effectively, and stabilize power systems as climate pressures rise.

Panel 7: AI for Global Sustainable Development: Technology, Policy, and Impact closed out the summit’s programming with a wide-angle view. 

This session considered how AI might help tackle massive global issues, from climate change to urban planning. Panelists emphasized the need for interdisciplinary solutions, where technology works with thoughtful policy and community engagement. 

While Day 1 offered a foundation in infrastructure and hardware, BU’s panels extended that framework to grids, governments, and the world.

With climate urgency mounting and AI capabilities expanding, the Green AI Summit marked a critical step in reshaping how innovation can serve the planet—not just businesses. For attendees, it wasn’t just a conversation but a call to action.

In closing the weekend’s events, Ayse Coskun, a professor of Electrical and Computer Engineering at BU and the Director of the Center for Information and Systems Engineering, expressed the importance of the collaboration between industry experts and academia. 

“We can use the support of AI; all of these things can only happen with the right groups together.”