Investor and Shark Tank star Kevin O’Leary is sounding the alarm on the rapid expansion of AI data centers, arguing that developers can no longer assume community acceptance without radical transparency. In a post on X and a follow-up CNN interview, O’Leary laid out a four-pillar framework: bring your own power, disclose water usage and emissions, create high-paying jobs, and prove the economic benefits to local stakeholders.
Transparency Is No Longer Optional
“Data center developers have to earn the trust of the communities where we build, and that starts with complete transparency,” O’Leary wrote on August 3, 2026. He emphasized that the era of tapping the local grid is over: “You can’t tap the grid. Otherwise, prices go up in the local community.” His own planned project in Box Elder, Utah, will publish water permits, power sources, job counts, and tax revenue projections publicly.
O’Leary dismissed concerns that AI facilities consume excessive water, stating that modern cooling systems recycle most of their supply and that “data centers today use as much water as a golf course does.” He endorsed journalist Joe Nocera’s call for greater industry transparency, saying “100 percent he’s right.”
Grid Constraints Drive On-Site Power Trend
The pressure on electrical infrastructure is already reshaping development strategies. Hyliion Holdings Corp CEO Thomas Healy noted that many operators are turning to on-site generation after failing to secure sufficient grid access, with interconnection delays becoming a major barrier to AI expansion. Meanwhile, Nvidia Corp CEO Jensen Huang suggested space-based computing could eventually process workloads closer to data sources, though cooling, launch costs, and infrastructure hurdles make widespread adoption years away.
Backlash Extends Beyond Energy
Billionaire entrepreneur Mark Cuban argued that community resistance reflects broader anxieties about job displacement and wealth inequality, not just energy consumption. He urged companies to improve communication and invest in local workforce support as AI transforms labor markets.
Why This Matters for Investors
- Regulatory risk: Communities demanding transparency may trigger stricter permitting and environmental reviews.
- Capital intensity: “Bring your own power” mandates increase upfront capex for developers.
- ESG scrutiny: Water usage and emissions disclosures will become standard ESG metrics for data center REITs and tech giants.
- Geographic diversification: Developers may favor regions with abundant renewable energy and supportive local policies.
FAQ
How much electricity do AI data centers actually consume?
Large-scale AI training clusters can draw 100–300 megawatts each—equivalent to a small city. The International Energy Agency estimates data centers could account for up to 4% of global electricity demand by 2030, driven largely by AI workloads.
What does “bring your own power” mean in practice?
It means developers must secure dedicated generation—often solar, wind, natural gas, or small modular nuclear—rather than relying on the local utility grid. This reduces strain on community infrastructure but raises project costs and complexity.
Are there investment vehicles tied to this trend?
Yes. Investors can access the theme through data center REITs (e.g., Digital Realty, Equinix), power infrastructure funds, and ETFs focused on AI hardware, renewable energy, or digital infrastructure. Private-market platforms like Fundrise and EquityMultiple also offer accredited investors direct exposure to real estate and credit strategies backing these facilities.
