Why Apple may be the one Big Tech company to hit its climate goals in the AI era
A decade ago, tech companies were seen as climate leaders, moving faster than the rest of corporate America in the transition to clean energy. Now, as soaring data center demand pushes them toward greater use of fossil fuels, these same companies are increasingly cast as climate villains.
One analysis of 60 of the largest planned data centers in the U.S. found that they could generate more than 100 million metric tons of CO2 emissions annually, roughly equivalent to the emissions from 24 million gas-powered cars. And as some tech companies turn to natural gas to power their data centers, they’re beginning to backtrack on their climate commitments.
Apple has taken a different approach. The company has rolled out AI more gradually than many of its peers and largely avoided the massive data center buildout driving the industry’s rise in emissions. Some of its AI processing happens on-device, using technology that Apple designed to boost efficiency, rather than in data centers.
When Apple uses the cloud, all of that compute runs on renewable energy, both at Apple’s own data centers—which have sourced wind and solar power since 2014—and at the third-party data centers it also uses. Apple’s carbon footprint is still dominated by its products and supply chain, not data centers, and those emissions are falling.
The company has a goal to cut emissions 75% by 2030, compared to a baseline in 2015, and will cover the remaining gap with carbon credits. So far, it reports that emissions have fallen 60%. As Apple’s AI use grows, some environmental advocates have questioned what path the company will take. But when Fast Company asked Apple if growing AI use threatens its climate goals, the company said no.
“We are on track towards our ambitious goal to be carbon neutral,” says Sarah Chandler, VP of environment and supply chain innovation at Apple. “That’s for our entire carbon footprint and by 2030. So, no changes there.”
At the same time that Apple cut emissions 60%, its revenue grew more than 75%. “We’re seeing revenue grow and carbon come down,” Chandler confirms. “And I think that’s really proof that you absolutely can lower your impact on the planet while growing your business.”
Apple’s climate challenge starts with its products
The biggest chunk of Apple’s emissions comes from its products, not data centers. Last year, the company’s total emissions hit 15.2 million metric tons of CO2; roughly half of that came from manufacturing, and more than a quarter came from the energy used by customers plugging in iPhones and Apple watches. Another 16% came from shipping products.
The company has been making progress. That includes a major shift to recycled materials—the MacBook Neo, for example, uses everything from a recycled aluminum enclosure to recycled gold in circuit boards and recycled cobalt in the battery. (Overall, the laptop is made from 60% recycled content.) Products like the new iPhone Duo are designed to be as durable and efficient as possible.
Using recycled materials has benefits beyond sustainability. Apple is now in a better position than many of its peers to deal with supply chain shortages for some critical materials. “A relatively small investment in recycling innovation and technology, done well, unlocks an incredible amount of material,” Chandler says.
The company has also pushed its suppliers to source clean energy. “Apple’s shown a much greater commitment to increasing the share of renewable energy used to manufacture its products, including at a time when demand for very energy-intensive chips has been going through the roof as well,” says Rachel Kitchin, senior corporate climate campaigner with the nonprofit Stand.earth.
Some of the company’s work has been controversial. When it cut the carbon footprint of Apple Watches by 75% and then bought carbon credits to cover the rest of the carbon footprint, it labeled the watches “carbon neutral.” A lawsuit in Germany argued that was misleading because the carbon credits involved short-term support for forests; a court found that the plan didn’t offset the long-term impact of emissions. Apple removed the labels, though it argued that it was losing an important way to communicate genuine progress.
The company’s 2030 goal takes the same approach for its entire supply chain—reducing emissions 75% and covering the remainder with nature-based carbon credits. By 2050, it plans to reduce emissions by 90%.
The company still needs to go further now, say environmental advocates. “They’re making progress, but there’s still a lot more to be done,” Kitchin says. “We would like to see 24/7, around-the-clock renewable energy targets.”
Currently, when companies set targets for 100% renewable electricity, she says, that’s typically calculated over a year—meaning that some of the time energy is actually coming from fossil fuels. A 24/7 goal would mean that operations relied on clean energy and battery storage all the time.
As the clock ticks down to 2030, Chandler says it’s getting more difficult to solve the remaining challenges. “We’ve [already done] a lot of the easier stuff,” she says. “From here, it may not be exactly linear progress. It’s going to be really tougher unlocks, I think, that are going to require an awful lot of collaboration and an awful lot of thinking about system change instead of just like one little piece of the puzzle.”
But the belief is that there’s a line of sight to meeting the 2030 goal. As the company works toward its much more ambitious 2050 goal, collaboration will be even more important. The transportation system, for example, needs to change across many players for Apple to be able to ship products without emissions. “Those are not things one company can do by themselves,” Chandler says.
A different approach to AI
AI is a different challenge, and one where other tech companies have faltered. This summer, Meta pulled out of RE100—an initiative it joined a decade ago where companies commit to using 100% renewable energy—because its investments in gas power meant that it no longer qualified. (In one deal earlier this year, Meta committed to seven new gas plants for a sprawling data center in Louisiana.)
Microsoft, which recently signed a deal with Chevron for a new gas plant at a data center in Texas, saw its emissions jump up 25% last year. Google’s overall emissions grew 18% in 2025, and the company’s chief sustainability officer said that “reaching our climate moonshot is getting harder.”
Other tech companies are pushing for AI “essentially at all costs,” Kitchin says. “In contrast, Apple is quite interesting because it actually does seem like it’s doing things differently. It’s not driving hard on infrastructure growth in the same way.” As AI use grows, she says, “this is definitely a crossroads moment for Apple’s future as a sustainable company.”
Apple’s approach differs in a couple of ways. First, the company has taken a more careful approach to using AI in the first place. When Apple rolled out Siri AI this summer, Craig Federighi, the company’s SVP of software engineering, threw some shade on competitors, arguing that some “appear to be racing forward, seemingly pursuing AI for the sake of AI, without clear regard for the people—all of us—that it’s ultimately meant to serve.”
By default, moving slower has meant less need for data centers. But the company is also using AI differently. Much of the AI on the new iPhone, for example, can be processed on the phone itself. That’s good for privacy reasons and is likely more efficient than sending each request to the cloud and running the servers, cooling, and other equipment at a data center. (Apple hasn’t shared specific numbers.)
And while it’s too early to say without data, it’s possible that if a large number of AI requests are handled on hundreds of millions of phones, that could mean a reduction in how many new data centers need to be built. Apple declined to share the percentage of AI that’s handled on-device versus in the cloud, but also noted that any cloud usage runs on renewable energy, including when the company uses data centers owned by third parties.
Apple’s own data centers are optimized to run as efficiently as possible. The company also pays for renewable electricity to cover the energy that customers use when charging their phones.
The company wouldn’t comment on any specific plans for AI rollout, but Chandler says Apple is taking the same approach to AI as the rest of its business.
“We’re embedding environment in everything that we do,” she says. “And that’s no different for AI, right? We’re really designing to consider the environmental impact from the beginning.”
Powering AI with clean energy
Tech companies face real challenges in trying to source clean energy for data centers. “A lot of it comes down to the mismatch between how policies, regulations, and utility business models are set up in this country and the technologies available today in 2026 to meet the moment that we’re in,” says Mark Dyson, managing director for electricity at the nonprofit RMI, which is working with the tech industry on the problem.
For example, virtual power plants—networks of devices like solar panels and batteries at homes that can be coordinated to support the grid—could play a role in helping data centers avoid building as many gas power plants. But even though some tech companies are beginning to use virtual power plants in limited ways, policies across most of the country make it hard to roll them out at the scale that’s needed. The utility business model, which pushes utilities to invest in expensive assets, is another challenge.
The process to add more power to the grid is also painfully slow. And while there are clean ways to add “behind the meter” or private power at data centers—like using solar and cheap used EV batteries—some companies have decided that it’s simpler to generate their own gas power instead. (That’s getting more challenging as some gas turbines face waitlists of up to seven years.)
Pollution is one of the reasons data centers are struggling to get support from communities—especially when the public is increasingly questioning whether we actually want widespread AI in the first place.
“I think one advantage that Apple has in terms of being a relatively slower mover in this AI boom is that they can learn from the lessons of how the hyperscalers have so far proceeded,” says Kelly Poole, a climate activist at the nonprofit As You Sow. “And the tech companies have received major backlash for changes to their emissions trajectories. Apple needs to be watching this backlash that these companies are getting, and should be warned that there’s a real benefit to maintaining your climate goals.”
Tech companies can help push new technologies forward. In Minnesota, Google helped pay for a new type of long-duration battery to support clean power on the grid. “That’s an example where [the utility] otherwise wouldn’t have taken a chance on this new battery technology,” Poole says.
Apple could also potentially play a role, she says, by bringing more funding to emerging tech and lobbying for new policies—such as clean transition tariffs—that would allow private companies to directly pay for new renewable energy to be added to the grids hosting their data centers.
Apple has eight company-owned data centers now, powered by renewable energy projects like wind and solar farms. The company is also expanding capacity in Iowa, Nevada, and Oregon. Chandler says the company doesn’t think that the availability of clean energy will be a constraint.
When it comes to AI, it’s not clear that others can copy Apple’s approach to reducing its environmental impact. (Obviously, from a climate perspective, the entire industry needs to find solutions, not just one company.) But Apple’s experience points to a broader question: whether companies can maintain long-term climate commitments as their businesses evolve.
“The renewable energy in our supply chain, the recycled content that we’re shipping, this stuff absolutely can work, and it’s not a trade-off,” Chandler contends. “It doesn’t have to come at the expense of your business growing.”
