The AI Industry, Data Center Buildout, and How to Take Power Back
Why It Matters
The data‑center surge jeopardizes environmental health, equity, and democratic oversight; coordinated policy action offers a tangible path to curb this unchecked expansion.
Key Takeaways
- •AI industry fuels rapid data‑center construction, straining resources.
- •3,000 data centers planned for 2025 threaten water, grid, health.
- •Concentrated tech power embeds bias, reduces accountability in AI tools.
- •Northstar toolkit offers policy levers to halt or slow buildout.
- •Community coalitions can leverage local, state, federal actions for change.
Summary
The AI Now Institute and the Data Center Working Group hosted a webinar to expose how the AI industry’s appetite for compute is driving an unprecedented data‑center buildout and to outline concrete policy tools for communities to reclaim control.
Speakers highlighted that roughly 3,000 data centers were under construction or planned for 2025, draining water supplies, inflating energy bills, destabilizing the grid, and worsening air quality. The industry projects up to $2 trillion in annual revenue, a trajectory that threatens job displacement, environmental degradation, and the concentration of wealth in a handful of tech giants.
Ama Kak, former FTC AI adviser, warned that AI concentrates power both at the point of deployment—embedding bias and obscuring accountability—and upstream, where a few firms control models, data, and infrastructure. She cited a Michigan case where an AI system falsely flagged 93 % of unemployment fraud claims, illustrating how “neutral” technology can perpetuate austerity.
The session introduced the Northstar data‑center policy toolkit, a set of zoning, water‑use, and taxation levers designed to stop or slow the buildout at local, state, and federal levels. By mobilizing community coalitions and leveraging these policy interventions, organizers aim to redirect tech infrastructure toward the public good rather than corporate profit.
Comments
Want to join the conversation?
Loading comments...