Dive Brief:
- The Biden administration convened leaders from artificial intelligence companies, hyperscalers, data center operators and utilities on Sept. 12 to strategize how to meet the country’s infrastructure needs for AI projects, per a White House news release.
- AI requires abundant amounts of electricity, water and land for data centers as well as factories to make semiconductors — not to mention workers to build and staff it all. Attendees discussed strategies to meet clean energy, permitting and labor requirements for developing large-scale data centers and power infrastructure needed for advanced AI.
- The Biden administration also announced a new interagency task force to coordinate federal policies to make sure they advance AI data center development in line with the White House’s economic, national security and environmental goals. It is also upping assistance to agencies that handle data center permitting.
Dive Insight:
The rise of AI has big implications for the building industry: The seemingly amorphous technology relies on enormous amounts of physical infrastructure and natural resources. The trend is driving demand for AI-ready data centers, even while many customers express wariness about AI.
The rapid growth of the energy-intensive tech has sparked concern about AI straining the country’s aging power grid. An average ChatGPT query uses nearly 10 times as much electricity to process as a Google search, and Goldman Sachs estimates that data center power demand will balloon 160% by 2030 to accommodate AI. At the same time, the carbon dioxide emissions of data centers may more than double between 2022 and 2030, per Goldman Sachs.
That surge in power demand could derail the Biden administration’s goal to decarbonize the power sector by 2035 to mitigate climate change. That’s why the White House wants to ensure clean energy powers future AI infrastructure, according to the release.
Discussing these challenges last week were tech titans like OpenAI CEO Sam Altman, Amazon Web Services CEO Matt Garman, Nvidia Corp. CEO Jensen Huang and President and CIO of Alphabet Ruth Porat. OpenAI alone wants to spend tens of billions on AI infrastructure in the U.S., such as data centers, energy capacity and transmission and semiconductor manufacturing, according to Data Center Knowledge.
The Biden administration also wants to make sure that the AI boom creates work for Americans. Roundtable participants discussed ways to create good-paying jobs, including roles for pipefitters and electricians, per the release.
Biden has continued to push for responsible AI development with his 2023 Executive Order on AI and subsequent July actions, and these new initiatives advance this goal. According to the White House, new efforts include:
- AI Data Center Infrastructure task force: Led by the National Economic Council, National Security Council and the White House Deputy Chief of Staff’s office, it will identify existing authorities and areas where legislative action is needed to support AI data center development, among other actions.
- Permitting help: Besides upping technical assistance to federal, state and local authorities that handle data center permitting, the Permitting Council will work with AI data center developers to set timelines for federal agency action. It will also allocate funds to agencies that accelerate evaluations for FAST-41 covered clean energy projects that support data centers.
- Nationwide permits: The U.S. Army Corps of Engineers will identify nationwide permits that can help expedite the construction of eligible AI data centers, and share that information with AI data center developers to expedite projects.
- DOE AI data center engagement team: The Energy Department is creating a team to leverage programs to support AI data center development. DOE has a suite of resources — including loans, grants, tax credits and technical assistance — to help data center owners and operators secure clean, reliable energy solutions. The agency also plans to host a series of convenings with members of the industry to generate solutions.
