I joined AD of AI Governance Lead as a guest for a live discussion on the environmental and economic footprint of AI data center buildout, grounded in what's happening in my own region. New Jersey electric rates have risen 30 to 50% in some areas, and the conversation traces that increase to PJM, the regional grid operator that sets rates across a corridor running from New Jersey through eastern Pennsylvania to DC, where data center demand has become a dominant driver of new grid costs passed on to ratepayers regardless of their own usage.

The conversation also covers the water side of the impact, which gets less attention than energy. Data center cooling draws heavily on fresh water that can't be returned to the system once used, and some communities near large facilities have reported water restrictions as a direct result. We discuss the proposed 1.3 million square foot data center at the former Pennhurst site outside Philadelphia, a project currently delayed by community pushback, alongside a broader pattern: data centers create relatively few permanent local jobs, most of the economic benefit accrues to the corporations building them, and the states and municipalities hosting these facilities have little public visibility into actual water and energy usage because disclosure isn't mandated.

Host: AD, host of the AI Governance, Ethics and Leadership Substack and podcast

Links: Watch on Substack

Skills demonstrated: AI infrastructure policy analysis, public utility and grid regulation, environmental impact communication, translating regulatory detail (PJM, state legislation) for public audiences, live public-facing commentary

Relevant context: Data center energy and water demand has become one of the least visible cost drivers in AI policy debates, often overshadowed by discussion of consumer-facing AI tools. This conversation focuses specifically on the regional grid and municipal water systems affected by buildout in the New Jersey/Philadelphia corridor, an underexamined angle even within AI governance coverage.