Environmental activist Erin Brockovich has launched a campaign demanding greater transparency from technology companies over the environmental impact of AI data centres, according to TechCrunch AI. The effort targets what Brockovich characterises as systematic corporate secrecy around water consumption, energy use, and emissions from rapidly expanding AI infrastructure.
The campaign arrives as hyperscale data centre construction accelerates to meet generative AI computing demands. Brockovich, whose legal work against Pacific Gas & Electric inspired the 2000 film bearing her name, is pressing for mandatory disclosure requirements that would force operators to reveal facility-level environmental metrics currently treated as proprietary information.
According to the TechCrunch report, Brockovich’s organisation is specifically challenging the practice of aggregating environmental data at corporate level rather than disclosing it per facility. This opacity, she argues, prevents communities from understanding local impacts on water tables, electrical grids, and air quality from data centres that can consume millions of litres of water daily for cooling systems.
The timing proves significant as multiple jurisdictions consider new regulations. The European Union’s Energy Efficiency Directive already requires large data centres to report energy consumption and waste heat, whilst several US states are examining similar measures. Ireland, home to numerous hyperscale facilities, has imposed a de facto moratorium on new data centre connections to Dublin’s electrical grid due to capacity concerns.
Tech companies have largely resisted facility-specific disclosure, citing competitive sensitivity and security concerns. Major operators including Microsoft, Google, and Amazon Web Services publish sustainability reports with company-wide metrics but rarely break down consumption by individual sites. Industry groups argue that aggregated reporting provides sufficient transparency whilst protecting operational details.
Business Impact
Increased disclosure requirements would create immediate compliance costs for data centre operators, particularly smaller players lacking dedicated environmental reporting infrastructure. However, companies with strong environmental performance metrics could gain competitive advantage in enterprise sales, where corporate sustainability commitments increasingly influence procurement decisions.
Real estate investment trusts specialising in data centre properties face potential valuation pressure if mandatory disclosures reveal higher-than-expected environmental footprints. Conversely, operators investing in liquid cooling, renewable energy, and water recycling technologies could differentiate themselves in a market where environmental impact becomes quantifiable and comparable.
The campaign also affects AI companies indirectly. Training large language models requires enormous computational resources, with some estimates suggesting GPT-3 training consumed 700,000 litres of water for cooling. If data centre operators face stricter disclosure rules, AI developers may encounter pressure to report the environmental cost of model training and inference.
Regulatory Momentum
Brockovich’s advocacy follows growing regulatory attention to data centre environmental impacts. California’s State Water Resources Control Board has begun requiring water use reporting from large facilities, whilst Virginia—home to the world’s largest data centre concentration—is considering similar measures after community opposition to new projects.
The campaign has attracted support from environmental groups and some local government officials in data centre hubs. However, economic development agencies often resist additional requirements, viewing data centres as sources of tax revenue and employment despite their relatively small workforce per facility.
Industry observers expect the debate to intensify as AI adoption drives further infrastructure expansion. Goldman Sachs projects data centre power consumption could grow 160% by 2030, predominantly driven by AI workloads. Without corresponding transparency measures, communities may face environmental impacts they cannot quantify or challenge.
The immediate focus will be whether Brockovich’s campaign gains traction in state legislatures currently considering data centre regulations. California, Oregon, and Washington—all significant data centre markets—have active environmental disclosure bills that could incorporate facility-level requirements. Success in any major jurisdiction would likely establish a de facto national standard, as operators would face pressure to provide consistent disclosure across markets.
The campaign represents a test of whether environmental activism can force transparency in an industry that has largely avoided the facility-level scrutiny applied to traditional industrial operations, potentially establishing precedent for how AI infrastructure expansion is monitored and regulated.
