Tech industry titans meet with White House officials to discuss AI energy needs

Sept. 12 (UPI) -- Key Biden Administration personnel met with leaders of the biggest technology companies at the White House on Thursday to discuss new infrastructure needed to support artificial intelligence, according to multiple reports.

Leaders of tech industry titans OpenAI, Anthropic, Nvidia, Microsoft and Google met with Cabinet members such as Commerce Secretary Gina Raimondo and Energy Secretary Jennifer Granholm to discuss the need for more AI data centers and the beefed-up energy infrastructure that will be needed to supply them, according to White House sources cited by CNN , CNBC and The Hill .

Among those in attendance were Nvidia CEO Jensen Huang, OpenAI CEO Sam Altman, Google Chief Investment Officer Ruth Porat and Anthropic CEO Dario Amodei, the sources said.

Reflecting the importance AI infrastructure needs hold for national security, the environment and many other sectors, the meeting also reportedly included U.S. National Security Adviser Jake Sullivan, White House climate advisers Ali Zaidi and John Podesta and White House chief of staff Jeff Zients .

The unannounced meeting came as AI has become the biggest driver of technology businesses around the world and one the prime focuses of Wall Street investors.

As AI is incorporated into more and more applications, concerns about its energy-hungry nature and the ability of U.S. utilities supply the growing number of data centers needed to power them have multiplied.

Investment bankers Goldman Sachs reported in May that the AI revolution is expected to boost data center power demand by 160% within the next six years.

At present, data centers worldwide consume 1% to-2% of overall power, but this percentage will likely rise to 3% to 4% by the end of the decade, driving a binge of electricity growth unlike anything seen in a generation.

Along the way, the carbon dioxide emissions of data centers likely will more than double between 2022 and 2030, they estimated. That's because a single ChatGPT query requires 2.9 watt-hours of electricity, compared with 0.3 watt-hours for a Google search, according to a study by the Electric Power Research Institute.

The rapid growth of AI power demand is highlighting the need for data center efficiency improvements and increased flexibility, closer coordination between data center developers and electric companies and the need for beefed-up investments in power grids so that AI can be accommodated "without negatively impacting other customers," the study notes.

Goldman Sachs estimated the overall increase in data center power consumption from AI to be on the order of 200 terawatt-hours per year between 2023 and 2030, eventually comprising about 19% of all data center power demand.

"President [Joe] Biden and Vice President [Kamala] Harris are committed to deepening U.S. leadership in AI by ensuring data centers are built in the United States while ensuring the technology is developed responsibly," a White House spokesperson said in a statement issued to media outlets.