CALIFORNIA STATE – California’s artificial intelligence boom is not just happening inside San Francisco offices or Silicon Valley startups. It is also showing up in city planning files, environmental reviews, utility requests, military energy projects and neighborhood fights over land, power, water and diesel backup generators.
As of July 2026, California had 45 operating data centers totaling about 899 megawatts of capacity, with another 27 planned projects that could add roughly 1,773 megawatts, according to data-center tracking firm Cleanview.
Not every project below is at the same stage. Some are approved. Some are under environmental review. Some are already under construction. One is in political limbo. And not every company labels its project as an “AI data center.”
But all seven cities are dealing with the new reality of AI-era computing: massive buildings, huge electricity needs, and public pressure to explain exactly what communities are getting in return.
1. San Jose
San Jose is arguably the center of California’s next data center wave.
The city now keeps a public list of data center applications, including projects under environmental review and others that have already been approved. Current proposals include two four-story data center buildings totaling about 630,000 square feet at 2549 Orchard Parkway, a 206,250-square-foot project at 350 W. Trimble Road, and multiple downtown proposals that combine data center space with housing or other uses.
San Jose has also approved major data center projects, including a 99-megawatt campus at 4160 Zanker Road and 1657 Alviso-Milpitas Road, and a 91-megawatt project at 1849 Fortune Drive, 2000 Trade Zone Boulevard and 2400 Ringwood Avenue, according to the city’s data center project list.
The biggest headline may be Prologis, which has won approval for a roughly 396-megawatt campus with four two-story data center buildings totaling about 1.68 million square feet. Data Center Dynamics reported that Microsoft has also broken ground on a 48-megawatt Alviso data center campus, described as the company’s first self-owned and operated data center in San Jose.
San Jose officials are now working on uniform standards for data centers, including rules and best practices around energy use, water use, backup power, air quality, greenhouse gas emissions, noise and community notification.
2. Gilroy
Gilroy is becoming one of the Bay Area’s most closely watched data center battlegrounds.
Amazon Web Services is behind the Gilroy Data Center project at 8050 Camino Arroyo, on a 56-acre undeveloped parcel near Arroyo Circle. The project is planned in two phases, with two data center buildings totaling about 438,500 square feet, plus a security building, substation, switchyard, off-site PG&E upgrades and a possible future battery energy storage system of up to 50 megawatts.
The first phase calls for a roughly 218,000-square-foot building with a 49-megawatt PG&E connection and 25 emergency generators. City records show the project has moved beyond environmental review and into permitting, with grading and utilities permits issued in December 2025 and a building permit issued in March 2026.
The project has also drawn neighborhood pushback. NBC Bay Area reported in May 2026 that residents were objecting as construction activity became visible, with critics raising concerns about transparency, pollution and local impacts.
3. Pittsburg
In Contra Costa County, Pittsburg has approved one of the East Bay’s most significant data center proposals.
The AVAIO Perseus Data Center is planned for the former Delta View Golf Course property along West Leland Road, within the Pittsburg Technology Park area. The city describes the project as an approved data center on a 76-acre site intended to support advanced technology infrastructure and economic development.
AVAIO says the Pittsburg project at 2232 Golf Club Road is designed for AI and cloud-driven deployments, with an initial 99-megawatt grid connection and first occupancy expected in 2027. The company also says the project will use 100% recycled wastewater and more than 90% zero-carbon grid power.
A related California Energy Commission filing describes a 92-megawatt backup generating system with 37 diesel generators for emergency use, serving a three-story, 347,740-square-foot data center building.
The project has become controversial. KTVU reported in June 2026 that residents and environmental advocates have raised concerns about energy use, noise, water and traffic, while the city has argued that the project would use recycled water and that its power needs would be handled separately from residential PG&E customers.
4. Hayward
Hayward is another East Bay city with a major planned data center campus.
The SVY03A Data Center Campus, proposed by STACK Infrastructure, is planned at 26203 Production Avenue, near Production Avenue and Investment Boulevard. California Energy Commission records describe a project with a three-story data center building, a one-story data center building, backup generators, a security building, an on-site substation and a PG&E switching station.
Cleanview lists the Hayward project as a planned 77-megawatt data center with about 332,000 square feet on an 11-acre site.
Hayward’s project has not received the same level of public attention as Pittsburg or Gilroy, but it shows how the data center buildout is spreading beyond the traditional Santa Clara County corridor and into industrial parts of the East Bay.
5. Sacramento
The Sacramento region is also becoming part of California’s AI infrastructure map.
Prime Data Centers announced in May 2026 that it had broken ground on SMF02, its second data center on a Sacramento-area campus. The project is planned as a 150,000-square-foot facility with 18 megawatts of critical IT load. Prime described the facility as part of growing demand for “AI-ready digital infrastructure” and said it is designed for hyperscale and high-density deployments.
Data Center Dynamics reported that SMF02 is located close to Prime’s existing facility at 2408 AK Street.
Compared with San Jose or Pittsburg, Sacramento’s project is smaller. But the AI angle is unusually explicit: Prime is directly tying the new buildout to high-density AI and high-performance computing workloads.
6. Lemoore
Not all of California’s AI data center projects are conventional city planning cases.
At Naval Air Station Lemoore, Ameresco is working with the U.S. Navy and CyrusOne on infrastructure and energy supplies for a 100-megawatt, AI-optimized data center. Ameresco describes the project as pairing an AI data center with critical energy infrastructure at the naval air station.
Data Center Dynamics reported that the project is expected to colocate the 100-megawatt data center with an on-site microgrid, giving the facility a dedicated energy strategy rather than relying only on the surrounding grid.
Because this project is on a military installation, it is not the same kind of local land-use fight as a downtown San Jose tower or a former golf course in Pittsburg. But it is a major example of how AI infrastructure is being linked with energy security, federal land and large-scale power planning.
7. Imperial
Imperial may be the most politically complicated project on this list.
The proposed Imperial Valley AI Data Center has been described as a nearly one-million-square-foot facility designed for advanced artificial intelligence and machine learning operations. CalMatters reported that the project has raised questions over water, energy and local decision-making, while Cleanview lists it at 330 megawatts and about 950,000 square feet.
Data Center Dynamics reported that the project, tied to Imperial Valley Computer Manufacturing, has raised concerns over transparency, water and local infrastructure, with the city of Imperial and Imperial County clashing over how the project was moving forward.
CalMatters also reported that county supervisors initially approved a key land step for the project, then later walked back that decision, declared a 45-day moratorium and formed a commission to study data center development.
Why These Projects Are Suddenly Everywhere
The AI boom is taking over California’s economy, and today’s powerful AI models need lots of computing power to operate.
Supporters say these projects can bring construction jobs, tax revenue, upgraded power infrastructure and a foothold in the next generation of cloud computing. Critics argue that there can be local impacts like diesel backup generators, noise, transmission lines, water demand, grid stress and huge buildings.
AI is an essential technology for the modern world, but building the infrastructure behind it can be a fraught process. For California cities, the next phase of AI may not look like a chatbot. It may look like a planning hearing.
