Back to briefings
Download PDF

April 26, 2026

Bain Capital seeks $2bn for 40% Bridge Data Centres Bridge Data Centres targets 3GW capacity by 2030 Energy Vault sees AI-driven storage demand; 10.5-year PG&E contract UC Riverside: 519ml water per 100-word AI search

Bain Capital kicking off a process to sell a chunky slice of Bridge Data Centres is a clean signal that the market for scaled Asian data centre platforms is still open—if you’ve got the right story and enough megawatts. The reported ask is about $2bn for at least 40%, implying a $5bn valuation for a company founded in 2017 and now talking about 3GW by 2030. In a week where “AI demand” shows up in everything from energy-storage order books to campus debates about water, the money is still chasing capacity—but the scrutiny is rising too.

The Big Stories

Bain Capital to sell 40% stake in Bridge Data Centres is the day’s headline for anyone tracking platform liquidity. Bain has reportedly started a sales process to offload at least 40% of the Singapore-based operator for about $2bn, valuing Bridge at around $5bn (S$6.4bn). Bridge says it’s targeting 3GW by 2030, spanning hyperscale build-to-suit and colocation—exactly the sort of “scale + pipeline” mix large infrastructure and sovereign pools want exposure to. The tell here is pricing: if $5bn is real (and financeable), it’ll reset expectations for other Southeast Asia platforms that have been hoping the exit window would reopen.

Energy Vault CEO: High oil prices could boost renewables puts a sharper edge on the power conversation: Robert Piconi says high oil prices and electricity volatility are driving a “sharp increase” in enquiries and firm orders for storage, with AI and data-centre electricity demand a direct driver. Energy Vault now owns and operates two plants, with two more under construction, and has a 10.5-year PG&E contract to provide up to 48 hours of green-hydrogen backup power. The subtext: as grid constraints tighten, storage and long-duration backup stop being “nice to have” and start looking like the gating items that determine whether new data centre capacity can actually run at high utilisation.

The industry’s resource footprint is also getting more specific in public debate. At Gannon University, a student/staff roundtable surfaced anxieties about data centres’ energy and water use, citing UC Riverside research that every 100-word AI search uses roughly 519 milliliters of water (Gannon students and staff discuss AI’s environmental impact concerns). Whatever you think of the exact methodology, the direction of travel matters: the “AI is intangible” narrative is being replaced by a more physical one—water, wastewater, power, and the local politics that come with them.

Behind the Headlines

Local permitting risk is quietly becoming a recurring theme, and the most telling line is often “more studies.” In Yukon, city officials confirmed multiple environmental studies are underway or planned to assess wastewater and other impacts ahead of a proposed data centre, with no financial or timeline details disclosed (Major environmental studies underway ahead of proposed data center). For investors, that absence of numbers is the point: uncertainty sits on the critical path, and “environmental studies” can mean anything from routine box-ticking to a years-long slog depending on the findings and local politics. When projects stall, it’s rarely because the demand case disappeared—it’s because the development process hit a constraint that spreadsheets don’t model well.

A separate commentary piece captures the broader mood shift: communities are increasingly bundling data centre growth into a wider argument about land, resource governance, and institutional capacity. Liz Pinkey argues for developing technology that protects the environment while enabling connectivity, while warning about data centre construction showing up at local advisory boards alongside broader policy changes that threaten local resources (Life With Liz: Develop technology to help save environment). She points to concerns like the U.S. Forest Service being “systemically dismantled” and proposals to lift mining restrictions in Minnesota’s Boundary Waters—different issues, but the same political ecosystem. The implication for operators is uncomfortable but straightforward: your project may be judged less on your own mitigation plan and more on whether residents trust the system around you.

Subscribe to Data Centres Briefings

Get AI-powered briefings delivered to your inbox

Region