Kevin O'Leary Blames Chinese Propaganda for U.S. Data‑Center Backlash – Industry and Administration Echo Claims

Announcement

Shark‑Tank billionaire Kevin O'Leary announced on May 10 that a $100 billion, 40,000‑acre data‑center project slated for Utah is being sabotaged by “hundreds of millions of dollars” in Chinese‑origin funding. In interviews with Fox News and subsequent posts on X, O'Leary claimed that the protestors were largely “bussed in” and that the campaign was part of a coordinated effort to curb U.S. dominance in artificial‑intelligence (AI) compute.

Interior Secretary Doug Burgum echoed the sentiment, describing the opposition as “not organic and local” and suggesting that “foreign‑directed propaganda” was targeting every new data‑center proposal across the country. The Washington Post’s exposé cited unnamed sources within the administration but offered no hard evidence linking Chinese money to the Utah protests.

{{IMAGE:2}}

---

Technical specs and supply‑chain context

Power demand and node density

The Utah project plans to host up to 2 GW of compute capacity, a figure comparable to the combined load of three large‑scale hyperscale campuses in the Pacific Northwest. At a typical 7 nm or 5 nm process node, each GPU or ASIC consumes roughly 300–350 W under full AI load. Assuming an average of 2 kW per rack, the facility would need about 1 000 000 kW of continuous power – a demand that would strain the regional grid unless paired with dedicated renewable contracts.

Water usage

Water‑cooled chillers for high‑density racks can draw 1.5–2 L per kW‑hour. For a 2 GW installation, annual water consumption can exceed 30 million cubic meters, roughly the yearly supply of a midsize city. Local municipalities have flagged such withdrawals as a risk to potable water supplies, especially during drought periods common in the Intermountain West.

Chip supply constraints

The AI boom has already exposed bottlenecks in advanced‑node silicon. Global foundries operating at 5 nm and 3 nm (e.g., TSMC, Samsung) are running at 85 % capacity, with lead times of 12–18 months for new wafer slots. The Utah data‑center’s projected demand for next‑generation GPUs could require upwards of 15 % of the annual 5 nm GPU allocation, effectively diverting chips from other customers and inflating prices for enterprise buyers.

Energy‑price impact

Historical data from the Electric Reliability Council of Texas (ERCOT) shows that each additional gigawatt of load can lift wholesale electricity prices by 3–5 cents per kWh during peak demand. Extrapolating to a 2 GW load, the local market could see a 6–10 cents/kWh increase, translating to a 15–25 % rise in residential electricity bills for nearby communities.

---

Market implications

Political risk premium

If the foreign‑interference narrative gains traction, developers may face higher permitting fees and longer review cycles. A 2023 study by the Brookings Institution estimated that political risk can add 5–10 % to total project capex for large‑scale infrastructure. For a $100 billion data‑center, that translates to an extra $5–10 billion in costs.

Supply‑chain ripple effects

The projected chip demand from the Utah campus could accelerate the already‑tight allocation of 5 nm and 3 nm wafers. OEMs that rely on the same silicon for consumer laptops and servers may experience longer lead times and higher unit costs, potentially slowing the rollout of next‑generation AI‑enabled devices.

Investor sentiment

While O'Leary’s claims have not been substantiated, the mere association with foreign meddling can sway institutional investors. ESG‑focused funds, which now account for roughly 30 % of U.S. equity assets, have begun to screen projects for geopolitical risk. A data‑center perceived as a target of foreign propaganda could see a discount of 2–4 % in valuation multiples compared with peers in less contentious regions.

Competitive dynamics

China’s own AI compute capacity is expanding at an estimated 20 % CAGR, driven by state‑backed cloud providers and aggressive subsidies for advanced‑node fabs. If U.S. developers encounter higher regulatory hurdles, Chinese firms could capture a larger share of the global AI‑training market, which is projected to exceed $150 billion by 2030.

---

Conclusion

The Utah data‑center controversy sits at the intersection of technology, resource management, and geopolitics. While O'Leary and certain officials point to Chinese propaganda as the primary catalyst, the underlying technical drivers—massive power draw, water consumption, and a strained advanced‑node supply chain—provide a concrete explanation for local opposition. Until verifiable evidence of foreign financing emerges, analysts should treat the foreign‑interference claim as a political narrative that adds a risk premium, rather than the root cause of community pushback.

---

For further reading on AI‑compute demand and chip supply constraints, see the latest reports from the Semiconductor Industry Association and the International Energy Agency.