Your phone's AI assistant is quietly helping overload the US power grid
Google, for example, is using lots and lots of electricity for its AI projects. Lots and lots!
Google has agreed to cut back electricity use at some of its data centers during times of high demand on the power grid. Yup, artificial intelligence has become very energy-hungry.
For Google, this marks the first time it has committed to temporarily limiting its AI workloads in coordination with utility providers. While these cutbacks only affect a small slice of power usage overall, they point to a growing tension between the rapid growth of AI and the limits of the energy infrastructure we rely on for all other things.
While Google didn't name the specific AI projects that are so power-demanding, I'm sure that it has to do with a thing smartphone users will be familiar with, namely – Gemini. For those who still haven't heard of it, that's the Big G's advanced AI assistant, which powers features like smart replies, email summaries, and photo editing across Android and Google apps. Although these tools seem simple on the surface, they are complex.
Data centers (that are being used by Google and other companies to develop and train AI) are now drawing so much electricity that some regions are struggling to keep up and Google has agreed to scale its effort down.
Utilities like Indiana Michigan Power and the Tennessee Valley Authority, which are part of the new agreements with Google, have reported rising demand from tech companies building out AI infrastructure. In certain areas, energy requests from these facilities are approaching or even exceeding the total available supply – raising concerns about power shortages and higher costs for homes and businesses.
To help prevent that, Google has agreed to slow down some of its AI-related computing when the grid is under stress. These types of arrangements, known as demand-response programs, are common in energy-heavy industries like manufacturing or crypto mining.
Smart features on our phones – now powered by systems similar to Gemini – come with an invisible footprint. The convenience of AI isn't just about speed or accuracy anymore. It's also starting to have real-world effects on things like electricity demand, power bills, and even grid reliability.
I pity those who don't feel a need for AI on their phones, yet experience power grid disruptions because of AI. Such irony.
For Google, this marks the first time it has committed to temporarily limiting its AI workloads in coordination with utility providers. While these cutbacks only affect a small slice of power usage overall, they point to a growing tension between the rapid growth of AI and the limits of the energy infrastructure we rely on for all other things.
Utilities like Indiana Michigan Power and the Tennessee Valley Authority, which are part of the new agreements with Google, have reported rising demand from tech companies building out AI infrastructure. In certain areas, energy requests from these facilities are approaching or even exceeding the total available supply – raising concerns about power shortages and higher costs for homes and businesses.
To help prevent that, Google has agreed to slow down some of its AI-related computing when the grid is under stress. These types of arrangements, known as demand-response programs, are common in energy-heavy industries like manufacturing or crypto mining.
Data center demand flexibility is still in the early stages and will only be available at certain locations. There are limits to how flexible a given data center can be, since high levels of reliability are critical for services like Search and Maps, as well as Cloud customers in essential industries like healthcare. [...] Managing data center load growth will require a portfolio of solutions, including new generation and transmission investments, but demand response can play an important role. Looking forward, we remain committed to collaborating with system operators, utilities, and industry partners to capture AI's immense opportunity while supporting a clean, reliable, and affordable energy system for everyone.
– Google's statement on a Google blog post, August 2025
Smart features on our phones – now powered by systems similar to Gemini – come with an invisible footprint. The convenience of AI isn't just about speed or accuracy anymore. It's also starting to have real-world effects on things like electricity demand, power bills, and even grid reliability.
FCC OKs Cingular\'s purchase of AT&T Wireless
Things that are NOT allowed:
To help keep our community safe and free from spam, we apply temporary limits to newly created accounts: