Justine Calma | The Verge

Trump’s AI plan is a massive handout to gas and chemical companies

2025-07-29T10:23:46-04:00

The Trump administration put out its vision for AI infrastructure in the US last week. It’s a dream for the fossil fuel and chemical industries — and a nightmare for wind and solar energy and the environment.

An “AI Action Plan” and flurry of executive orders Donald Trump signed last week read like manifestos on making AI less “woke” and less regulated. They’re packed with head-spinning proposals to erode bedrock environmental protections in the US, on top of incentives for companies to build out new data centers, power plants, pipelines, and computer chip factories as fast as they can.

It’s a deregulation spree and a massive handout to fossil fuels, all in the name of AI.

What the AI plan “is really about” is “using unprecedented emergency powers to grant massive new exemptions for data centers and specifically fossil fuel infrastructure,” says Tyson Slocum, energy program director at the consumer advocacy group Public Citizen. “I think they have a genuine interest in accommodating Big Tech’s priorities. But it’s an opportunity to marry their priorities for Big Oil.”

“It’s an opportunity to marry their priorities for Big Oil.”

Data centers are notoriously energy-hungry and have already led to a surge of new gas projects meant to satiate rising demand. But many tech companies have sustainability commitments they’ve pledged to meet using renewable energy, and as wind and solar farms have generally grown cheaper and easier to build than fossil fuel power plants, they’ve become the fastest-growing sources of new electricity in the US. Now, Trump wants to turn that on its head.

He signed an executive order on July 23rd meant to “accelerat[e] federal permitting of data center infrastructure.” It tells the Secretary of Commerce to “launch an initiative to provide financial support” for data centers and related infrastructure projects. That could include loans, grants, and tax incentives for energy infrastructure — but not for solar and wind power. The executive order describes “covered components” as “natural gas turbines, coal power equipment, nuclear power equipment, geothermal power equipment” and any other electricity sources considered “dispatchable.” To be considered dispatchable, operators have to be able to ramp electricity generation up and down at will, so this excludes intermittent renewables like solar and wind power that naturally fluctuate with the weather and time of day.

Trump’s AI planning document similarly says the administration will prioritize deploying dispatchable power sources and that “we will continue to reject radical climate dogma.” Already, Trump has dealt killer blows to solar and wind projects by hiking up tariffs and cutting Biden-era tax credits for renewables. The AI executive order goes even further to entrench reliance on fossil fuels and make it harder for new data centers to run on solar and wind energy.

“Right now, you do not qualify for expedited treatment if your data center proposal has wind and solar. It is excluded from favorable treatment,” Slocum says. “So what’s the statement for the market? Don’t rely on wind and solar.”

That’s not just environmentally unfriendly, it’s inefficient — considering the current backlog for gas turbines and because fossil fuel plants are generally slower and more expensive to build than onshore wind and solar farms. “This is not an energy abundance agenda. This is an energy idiot agenda,” Slocum adds.

The Trump administration wants to speed things up by rewriting bedrock environmental laws. Trump, ever the disgruntled real estate mogul, has railed against environmental reviews he says take too long and cost too much. He has already worked to roll back dozens of environmental regulations since stepping into office. Now, the executive order directs the Environmental Protection Agency (EPA) to modify rules under the Clean Air Act, Clean Water Act, Superfund law, and Toxic Substances Control Act to expedite permitting for data center projects.

“That is horrifying … These [laws] protect our public health. They protect our children. They protect the air we breathe and the water we drink,” says Judith Barish, coalition director of CHIPS Communities United, a national coalition that includes labor and environmental groups.

“This is an energy idiot agenda.”

The coalition has come together to fight for protections for workers in the chip industry and nearby communities. Semiconductor manufacturing has a long history of leaching harmful chemicals and exposing employees to reproductive health toxins. Santa Clara, California, home of Silicon Valley, has more toxic Superfund sites than any other county in the US as a result. The coalition wants to keep history from repeating itself as the US tries to revive domestic chip manufacturing and dominate the AI market.

AI requires more powerful chips, and Trump’s executive order fast-tracking federal permitting for data center projects includes semiconductors and “semiconductor materials.” Barish says “a chip factory is a chemical factory” because of all the industrial solvents and other chemicals semiconductor manufacturers use. That includes “forever chemicals,” for which the Trump administration has started to loosen regulations on how much is allowed in drinking water. Companies including 3M and Dupont have faced a landslide of lawsuits over forever chemicals linked to cancer, reproductive risks, liver damage, and other health issues, and have subsequently made pledges to phase out or phase down the chemicals. Now, manufacturers are jumping on the opportunity to produce more forever chemicals to feed the AI craze.

Ironically, we could see data centers and related infrastructure popping up on polluted Superfund sites that Silicon Valley has already left in its tracks. Trump’s executive order directs the EPA to identify polluted Superfund and Brownfield sites that could be reused for new data center projects (and tells other agencies to scour military sites and federal lands for suitable locations).

Office buildings are already situated on or adjacent to old Superfund sites where cleanup is ongoing; Google workers were exposed to toxic vapors rising from a Superfund site below their office back in 2013. Since it can take decades to fully remediate a site, oversight is key. “For Superfund sites in particular, these are the most contaminated sites in the country, and it is important that there are comprehensive reviews both for the people who are going to be working on the sites, as well as for the people who surround them,” says Jennifer Liss Ohayon, a research scientist at the Silent Spring Institute who has studied the remediation of Superfund sites.

But Trump wants to erode oversight for new data center projects that receive federal support — adding “categorical exclusions” to typical National Environmental Policy Act assessments. Environmental reviews that do take place could also be limited by the sheer lack of people power at federal agencies the Trump administration has hacked to pieces, including the EPA.

“America needs new data centers, new semiconductor and chip manufacturing facilities, new power plants and transmission lines,” Trump said before signing his AI executive orders last week. “Under my leadership we’re going to get that job done and it’s going to be done with certainty and with environmental protection and all of the things we have to do to get it done properly.” Good luck.

Facebook ranks worst for online harassment, according to a global activist survey

2025-07-25T14:40:23-04:00

Activists around the world are calling attention to harassment they’ve faced on Meta’s platforms. More than 90 percent of land and environmental defenders surveyed by Global Witness, a nonprofit organization that also tracks the murders of environmental advocates, reported experiencing some kind of online abuse or harassment connected to their work. Facebook was the most-cited platform, followed by X, WhatsApp, and Instagram.

Global Witness and many of the activists it surveyed are calling on Meta and its peers to do more to address harassment and misinformation on their platforms. Left to fester, they fear that online attacks could fuel real-world risks to activists. Around 75 percent of people surveyed said they believed that online abuse they experienced corresponded to offline harm.

“Those stats really stayed with me. They were so much higher than we expected them to be,” Ava Lee, campaign strategy lead on digital threats at Global Witness, tells The Verge. That’s despite expecting a gloomy outcome based on prior anecdotal accounts. “It has kind of long been known that the experience of climate activists and environmental defenders online is pretty awful,” Lee says.

Left to fester, they fear that online attacks could fuel real-world risks

Global Witness surveyed more than 200 people between November 2024 and March of this year that it was able to reach through the same networks it taps when documenting the killings of land and environmental defenders. It found Meta-owned platforms to be “the most toxic.” Around 62 percent of participants said they encountered abuse on Facebook, 36 percent on WhatsApp, and 26 percent on Instagram.

That probably reflects how popular Meta’s platforms are around the world. Facebook has more than 3 billion active monthly users, more than a third of the global population. But Meta also abandoned its third-party fact-checking program in January, which critics warned could lead to more hate speech and disinformation. Meta moved to a crowdsourced approach to content moderation similar to X, where 37 percent of survey participants reported experiencing abuse.

In May, Meta reported a “small increase in the prevalence of bullying and harassment content” on Facebook as well as “a small increase in the prevalence of violent and graphic content” during the first quarter of 2025.

“That’s sort of the irony as well, of them moving towards this kind of free speech model, which actually we’re seeing that it’s silencing certain voices,” says Hannah Sharpe, a senior campaigner at Global Witness.

Fatrisia Ain leads a local collective of women in Sulawesi, Indonesia, where she says palm oil companies have seized farmers’ lands and contaminated a river local villagers used to be able to rely on for drinking water. Posts on Facebook have accused her of being a communist, a dangerous allegation in her country, she tells The Verge.

The practice of “red-tagging” — labeling any dissident voices as communists — has been used to target and criminalize activists in Southeast Asia. In one high-profile case, a prominent environmental activist in Indonesia was jailed under “anti-communism” laws after opposing a new gold mine.

Ain says she’s asked Facebook to take down several posts attacking her, without success. “They said it’s not dangerous, so they can’t take it down. It is dangerous. I hope that Meta would understand, in Indonesia, it’s dangerous,” Ain says.

Other posts have accused Ain of trying to defraud farmers and of having an affair with a married man, which she sees as attempts to discredit her that could wind up exposing her to more threats in the real world — which has already been hostile to her activism. “Women who are being the defenders for my own community are more vulnerable than men … more people harass you with so many things,” she says.

Nearly two-thirds of people who responded to the Global Witness survey said that they have feared for their safety, including Ain. She’s been physically targeted at protests against palm oil companies accused of failing to pay farmers, she tells The Verge. During a protest outside of a government office, men grabbed her butt and chest, she says. Now, when she leads protests, older women activists surround her to protect her as a security measure.

In the Global Witness survey, nearly a quarter of respondents said they’d been attacked on the basis of their sex. “There’s evidence of the way that women and women of color in particular in politics experience just vast amounts more hate than any other group,” Lee says. “Again, we’re seeing that play out when it comes to defenders … and the threats of sexual violence, and the impact that that is having on the mental health of lots of these defenders and their ability to feel safe.”

“We encourage people to use tools available on our platforms to help protect against bullying and harassment,” Meta spokesperson Tracy Clayton said in an email to The Verge, adding that the company is reviewing Facebook posts that targeted Ain. Meta also pointed to its “Hidden Words” feature that allows you to filter offensive direct messages and comments on your posts and its “Limits” feature that hides comments on your posts from users that don’t follow you.

Other companies mentioned in the report, including Google, TikTok, and X, did not provide on-the-record responses to inquiries from The Verge. Nor did a palm oil company Ain says has been operating on local farmers’ land without paying them, as they’re supposed to do under a mandated profit-sharing scheme.

Global Witness says there are concrete steps social media companies can take to address harassment on their platforms. That includes dedicating more resources to their content moderation systems, regularly reviewing these systems, and inviting public input on the process. Activists surveyed also reported that they think algorithms that boost polarizing content and the proliferation of bots on platforms make the problem worse.

“There are a number of choices that platforms could make,” Lee says. “Resourcing is a choice, and they could be putting more money into really good content moderation and really good trust and safety [initiatives] to improve things.”

Global Witness plans to put out its next report on the killings of land and environmental defenders in September. Its last such report found that at least 196 people were killed in 2023.

Breaking down Trump’s big gift to the AI industry

2025-07-25T11:50:33-04:00

US President Donald Trump displays an executive order, which is about AI, in Washington, DC, on July 23rd. | Photo: AFP via Getty Images

President Donald Trump’s plan to promote America’s AI dominance involves discouraging “woke AI,” slashing state and federal regulations, and laying the groundwork to rapidly expand AI development and adoption. Trump’s proposal, released on July 23rd, is a sweeping endorsement of the technology, full of guidance that ranges from specific executive actions to directions for future research.

Some of the new plan’s provisions (like promoting open-source AI) have garnered praise from organizations that are often broadly critical of Trump, but the loudest acclaim has come from tech and business groups, whose members stand to gain from fewer restrictions on AI. “The difference between the Trump administration and Biden’s is effectively night and day,” says Patrick Hedger, director of policy at tech industry group NetChoice. “The Biden administration did everything it could to command and control the fledgling but critical sector … The Trump AI Action Plan, by contrast, is focused on asking where the government can help the private sector, but otherwise, get out of the way.”

Others are far more ambivalent. Future of Life Institute, which led an Elon Musk-backed push for an AI pause in 2023, said it was heartened to see the Trump administration acknowledge serious risks, like bioweapons or cyberattacks, could be exacerbated by AI. “However, the White House must go much further to safeguard American families, workers, and lives,” says Anthony Aguirre, FLI’s executive director. “By continuing to rely on voluntary safety commitments from frontier AI corporations, it leaves the United States at risk of serious accidents, massive job losses, extreme concentrations of power, and the loss of human control. We know from experience that Big Tech promises alone are simply not enough.”

For now, here are the ways that Trump aims to promote AI.

‘Consider a state’s AI regulatory climate when making funding decisions’

Congress failed to pass a moratorium on states enforcing their own AI laws as part of a recent legislative package. But a version of that plan was resurrected in this document. “AI is far too important to smother in bureaucracy at this early stage, whether at the state or Federal level,” the plan says. “The Federal government should not allow AI-related Federal funding to be directed toward states with burdensome AI regulations that waste these funds, but should also not interfere with states’ rights to pass prudent laws that are not unduly restrictive to innovation.”

To do this, it suggests federal agencies that dole out “AI-related discretionary funding” should “limit funding if the state’s AI regulatory regimes may hinder the effectiveness of that funding or award.” It also suggests the Federal Communications Commission (FCC) “evaluate whether state AI regulations interfere with the agency’s ability to carry out its obligations and authorities under the Communications Act of 1934.”

The Trump administration also wants the Federal Trade Commission (FTC) to take a hard look at existing AI regulations and agreements to see what it can scale back. It recommends the agency reevaluate investigations launched during the Biden administration “to ensure that they do not advance theories of liability that unduly burden AI innovation,” and suggests it could throw out burdensome aspects of existing FTC agreements. Some AI-related actions taken during the Biden administration that the FTC might now reconsider include banning Rite Aid’s use of AI facial recognition that allegedly falsely identified shoplifters, and taking action against AI-related claims the agency previously found to be deceptive.

‘Our AI systems must be free from ideological bias’

Trump’s plan includes policies designed to help encode his preferred politics in the world of AI. He’s ordered a revision of the Biden-era National Institute of Standards and Technology (NIST) AI Risk Management Framework — a voluntary set of best practices for designing safe AI systems — removing “references to misinformation, Diversity, Equity, and Inclusion, and climate change.” (The words “misinformation” and “climate change” don’t actually appear in the framework, though misinformation is discussed in a supplementary file.)

In addition to that, a new executive order bans federal agencies from procuring what Trump deems “woke AI” or large language models “that sacrifice truthfulness and accuracy to ideological agendas,” including things like racial equity.

This section of the plan “seems to be motivated by a desire to control what information is available through AI tools and may propose actions that would violate the First Amendment,” says Kit Walsh, director of the Electronic Frontier Foundation (EEF). “The plan seeks to require that ‘the government only contracts with’ developers who meet the administration’s ideological criteria. While the government can choose to purchase only services that meet such criteria, it cannot require that developers refrain from also providing non-government users other services conveying other ideas.”

‘Establishing a dynamic, ‘try-first’ culture for AI’

The administration describes the slow uptake of AI tools across the economy, including in sensitive areas like healthcare, as a “bottleneck to harnessing AI’s full potential.” The plan describes this cautious approach as one fueled by “distrust or lack of understanding of the technology, a complex regulatory landscape, and a lack of clear governance and risk mitigation standards.” To promote the use of AI, the White House encourages a “‘try-first’ culture for AI across American industry.”

This includes creating domain-specific standards for adopting AI systems and measuring productivity increases, as well as regularly monitoring how US adoption of AI compares to international competitors. The White House also wants to integrate AI tools throughout the government itself, including by detailing staff with AI expertise at various agencies to other departments in need of that talent, training government employees on AI tools, and giving agencies ample access to AI models. The plan also specifically calls out the need to “aggressively adopt AI within its Armed Forces,” including by introducing AI curricula at military colleges and using AI to automate some work.

‘Retrain and help workers thrive’

All this AI adoption will profoundly change the demand for human labor, the plan says, likely eliminating or fundamentally changing some jobs. The plan acknowledges that the government will need to help workers prepare for this transition period by retraining people for more in-demand roles in the new economy and providing tax benefits for certain AI training courses.

On top of preparing to transition workers from traditional jobs that might be upended by AI, the plan discusses the need to train workers for the additional roles that might be created by it. Among the jobs that might be needed for this new reality are “electricians, advanced HVAC technicians, and a host of other high-paying occupations,” the plan says.

‘Create a supportive environment for open models’

The administration says it wants to “create a supportive environment for open models,” or AI models that allow users to modify the code that underpins them. Open models have certain “pros,” like being more accessible to startups and independent developers.

Groups like EFF and the Center for Democracy and Technology (CDT), which were critical of many other aspects of the plan, applauded this part. EFF’s Walsh called it a “positive proposal” to promote “the development of open models and making it possible for a wider range of people to participate in shaping AI research and development. If implemented well, this could lead to a greater diversity of viewpoints and values reflected in AI technologies, compared to a world where only the largest companies and agencies are able to develop AI.”

That said, there are also serious “cons” to the approach that the AI Action Plan didn’t seem to get into. For instance, the nature of open models makes them easier to trick and misalign for purposes like creating misinformation on a large scale, or chemical or biological weapons. It’s easier to get past built-in safeguards with such models, and it’s important to think critically about the tradeoffs before taking steps to drive open-source and open-weight model adoption at scale.

‘Expedite environmental permitting’

Trump signed an executive order on July 23rd meant to fast track permitting for data center projects. The EO directs the commerce secretary to “launch an initiative to provide financial support” that could include loans, grants, and tax incentives for data centers and related infrastructure projects.

Following a similar move by former President Joe Biden, Trump’s plan directs agencies to identify federal lands suitable for the “large-scale development” of data centers and power generation. The EO tells the Department of Defense to identify suitable sites on military installations and the Environmental Protection Agency (EPA) to identify polluted Superfund and Brownfield sites that could be reused for these projects.

The Trump administration is hellbent on dismantling environmental regulations, and the EO now directs the EPA to modify rules under the Clean Air Act, Clean Water Act, and Toxic Substances Control Act to expedite permitting for data center projects.

The EO and the AI plan, similar to a Biden-era proposal, direct agencies to create “categorical exclusions” for federally supported data center projects that would exclude them from detailed environmental reviews under the National Environmental Policy Act. And they argue for using new AI tools to speed environmental assessments and applying the “Fast-41 process” to data center projects to streamline federal permitting.

The Trump administration is basically using the AI arms race as an excuse to slash environmental regulations for data centers, energy infrastructure, and computer chip factories. Last week, the administration exempted coal-fired power plants and facilities that make chemicals for semiconductor manufacturing from Biden-era air pollution regulations.

‘Navigating the complex energy landscape of the 21st century’

The plan admits that AI is a big factor “increasing pressures on the [power] grid.” Electricity demand is rising for the first time in more than a decade in the US, thanks in large part to data centers — a trend that could trigger blackouts and raise Americans’ electricity bills. Trump’s AI plan lists some much-needed fixes to stabilize the grid, including upgrading power lines and managing how much electricity consumers use when demand spikes.

But the administration is saying that the US needs to generate more electricity to power AI just as it’s stopping renewable energy growth, which is like trying to win a race in a vehicle with no front wheels. It wants to meet growing demand with fossil fuels and nuclear energy. “We will continue to reject radical climate dogma,” the plan says. It argues for keeping existing, mostly fossil-fueled power plants online for longer and limiting environmental reviews to get data centers and new power plants online faster.

The lower cost of gas generation has been killing coal power plants for years, but now a shortage of gas turbines could stymie Trump’s plans. New nuclear technologies that tech companies are investing in for their data centers probably won’t be ready for commercial deployment until the 2030s at the earliest. Republicans, meanwhile, have passed legislation to hobble the solar and wind industries that have been the fastest-growing sources of new electricity in the US.

‘Prioritize fundamental advancements in AI interpretability’

The Trump administration accurately notes that while developers and engineers know how today’s advanced AI models work in a big-picture way, they “often cannot explain why a model produced a specific output. This can make it hard to predict the behavior of any specific AI system.” It’s aiming to fix that, at least when it comes to some high-stakes use cases.

The plan states that the lack of AI explainability and predictability can lead to issues in defense, national security, and “other applications where lives are at stake,” and it aims to promote “fundamental breakthroughs on these research problems.” The plan’s recommended policy actions include launching a tech development program led by the Defense Advanced Research Projects Agency to advance AI interpretability, control systems, and security. It also said the government should prioritize fundamental advancements in such areas in its upcoming National AI R&D Strategic Plan and, perhaps most specifically, that the DOD and other agencies should coordinate an AI hackathon to allow academics to test AI systems for transparency, effectiveness, and vulnerabilities.

It’s true that explainability and unpredictability are big issues with advanced AI. Elon Musk’s xAI, which recently scored a large-scale contract with the DOD, recently struggled to stop its Grok chatbot from spouting pro-Hitler takes — so what happens in a higher-stakes situation? But the government seems unwilling to slow down while this problem is addressed. The plan states that since “AI has the potential to transform both the warfighting and back-office operations of the DOD,” the US “must aggressively adopt AI within its Armed Forces if it is to maintain its global military preeminence.”

The plan also discusses how to better evaluate AI models for performance and reliability, like publishing guidelines for federal agencies to conduct their own AI system evaluations for compliance and other reasons. That’s something most industry leaders and activists support greatly, but it’s clear what the Trump administration has in mind will lack a lot of the elements they have been pushing for.

Evaluations likely will focus on efficiency and operations, according to the plan, and not instances of racism, sexism, bias, and downstream harms.

‘Give the courts and law enforcement the tools they need’

Courtrooms and AI tools mix in strange ways, from lawyers using hallucinated legal citations to an AI-generated appearance of a deceased victim. The plan says that “AI-generated media” like fake evidence “may present novel challenges to the legal system,” and it briefly recommends the Department of Justice and other agencies issue guidance on how to evaluate and deal with deepfakes in federal evidence rules.

‘Improving the financial market for compute’

Finally, the plan recommends creating new ways for the research and academic community to access AI models and compute. The way the industry works right now, many companies, and even academic institutions, can’t access or pay for the amount of compute they need on their own, and they often have to partner with hyperscalers — providers of large-scale cloud computing infrastructure, like Amazon, Google, and Microsoft — to access it.

The plan wants to fix that issue, saying that the US “has solved this problem before with other goods through financial markets, such as spot and forward markets for commodities.” It recommends collaborating with the private sector, as well as government departments and the National Science Foundation’s National AI Research Resource pilot to “accelerate the maturation of a healthy financial market for compute.” It didn’t offer any specifics or additional plans for that.

How Trump’s war on clean energy is making AI a bigger polluter

2025-07-28T19:43:15-04:00

Sen. Dave McCormick (R-PA) and President Donald Trump during the inaugural Pennsylvania Energy and Innovation Summit at Carnegie Mellon University in Pittsburgh on July 15. | Photo: Bloomberg via Getty Images

At an AI and fossil fuel lovefest in Pittsburgh, Pennsylvania last week, President Donald Trump — flanked by cabinet members and executives from major tech and energy giants like Google and ExxonMobil — said that “the most important man of the day” was Environmental Protection Agency head Lee Zeldin. “He’s gonna get you a permit for the largest electric producing plant in the world in about a week, would you say?” Trump said to chuckles in the audience. Later that week, the Trump administration exempted coal-fired power plants, facilities that make chemicals for semiconductor manufacturing, and certain other industrial sites from Biden-era air pollution regulations.

If Trump has his way, the next generation of data centers will run dirtier than the last. It isn’t enough to kill renewables and pave the way for more coal and gas plants to power energy-hungry AI data centers. Trump is also obsessed with tossing out environmental protections.

“It costs much more to do things environmentally clean,” Trump claimed in an interview with Joe Rogan in October 2024. Upon his appointment to head the EPA (or, rather, run it into the ground), Zeldin said that he would be focused on “unleash[ing] US energy dominance” and “mak[ing] America the AI capital of the world.” The EPA announced thousands of layoffs on on July 18th, gutting its research and development arm.

“It costs much more to do things environmentally clean.”

At the Pennsylvania Energy and Innovation Summit, Trump attempted to take credit for private investments totaling around $36 billion for data center projects and $56 million for new energy infrastructure. The ceremony itself was mostly pomp and circumstance, but it’s telling that the Trump administration says it wants to make Pennsylvania a new hub for AI data centers. It’s a swing state that Republicans are eager to move into their column, but it’s also a major coal and gas producer. Sitting atop a major gas reserve, fracking in Pennsylvania (as well as Texas) helped usher in the “shale revolution” in the 2000s that made the US the world’s leading gas producer.

That was supposed to start changing under former President Joe Biden’s direction. He set a goal for the US to get all its electricity from carbon pollution-free sources by 2035. And in 2022, he signed the Inflation Reduction Act, which was full of tax incentives to make it cheaper to build out new solar and wind farms, as well as other carbon-free energy sources. If it had stayed intact, the law was expected to reduce US greenhouse gas emissions by around 40 percent this decade.

The law came at a crucial time for tech companies, which were expanding data centers as the AI arms race picked up steam. Electricity demand in the US is rising for the first time in more than a decade, thanks in large part to energy-hungry data centers. Google, Amazon, Microsoft, Meta, and other tech giants all have their own climate goals, pledging to shrink their carbon footprints by supporting renewable energy projects.

But Trump is making it harder to build those projects in the US. Republicans voted to wind down Biden-era tax incentives for solar and wind energy in the big spending bill they passed this month. The bill will likely decrease electricity generation capacity in 2035 by 340 GW, according to one analysis, with the vast majority of losses coming from solar and wind farms that will no longer get built.

All these new data centers still need to get their electricity from somewhere. “They won’t be powered by wind,” Trump said during the summit, repeating misleading talking points about renewable energy that have become a cornerstone of new climate denial. He signed an executive order in April, directing the Commerce, Energy, and Interior Departments to study “where coal-powered infrastructure is available and suitable for supporting AI data centers.” Trump, backed by fossil fuel donors, campaigned on a promise to “drill, baby, drill” — a slogan that he doubled down on again at the event. He also referenced the Homer City Generating Station, an old coal plant that’s reopening as a gas plant that will power a new data center.

The deals announced at the summit include Enbridge investing $1 billion to expand its gas pipelines into Pennsylvania and Equinor spending $1.6 billion to “boost natural gas production at Equinor’s Pennsylvania facilities and explore opportunities to link gas to flexible power generation for data centers.”

“They won’t be powered by wind.”

Data centers are a “main driver” for a boom in new gas pipelines and power plants in the Southeast, according to a January report from the Institute for Energy Economics and Financial Analysis (IEEFA). The Southeast is home to “data center alley,” a hub in Virginia through which around 70 percent of the world’s internet traffic flows through. Even if AI models become more efficient over time, the amount of electricity they’re currently projected to demand could lock communities across the US into prolonged reliance on fossil fuels as utilities build out new gas infrastructure.

Zeldin’s job now is essentially to remove any regulatory hurdles that might slow down that growth. From his first day in office, “it was clear that EPA would have a major hand in permitting reform to cut down barriers that have acted as a roadblock so we can bolster the growth of AI,” as Zeldin wrote in a Fox News op-ed last week. “A company looking to build an industrial facility or a power plant should be able to build what it can before obtaining an emissions permit,” he added. And after moving to roll back pollution regulations for power plants, the Trump administration is now reportedly working on a rule that would undo the 2009 “endangerment finding” that allows the EPA to regulate greenhouse gas emissions under the Clean Air Act.

Zeldin also writes that when it comes to Clean Air Act permits for polluters it considers “minor emitters,” the EPA will only meet “minimum requirements for public participation.” An AI Action Plan that the White House dropped on July 23rd proposes creating new categorical exclusions for data center-related projects from the National Environmental Policy Act (NEPA), a sunshine law that mandates input from local communities on major federal projects. The plan directs agencies to identify federal lands for the “large-scale development” of data centers and power generation.

There are other factors at play that could derail Trump’s fossil-fueled agenda, including a backlog for gas turbines in high demand. Solar and wind farms are still generally faster to build and a more affordable source of new electricity than coal or gas, and we could see some developers rush to complete projects before Biden-era tax credits fully disappear. One early bright spot for renewables was the fact that data centers used to train AI are theoretically easier to build close to far-flung wind and solar projects. Unlike other data centers, they don’t need to be built near population centers to reduce latency. They could also theoretically time their operations to match the ebb and flow of electricity generation when the sun shines and winds blow.

But so far, things are shaping up differently in the real world. “It’s just a race to get connected as quickly as possible,” says Nathalie Limandibhratha, senior associate US power at BloombergNEF.

Data center developers are also concerned that if they build facilities specifically to train AI closer to renewable energy, they could be left with stranded assets down the road. They’d rather keep building data centers close to population centers where they can repurpose the facility for other uses if needed. They also get more bang for their buck running 24/7, so data centers are leaning toward around-the-clock electricity generation from gas and nuclear energy (and nuclear energy has more bipartisan support than other sources of carbon-free energy).

“There’s no question right now that AI is driving greater fossil fuel use in the United States and really setting us back in terms of climate change,” says Cathy Kunkel, an energy consultant at IEEFA. Tech giants Google and Amazon made announcements coinciding with the Pennsylvania summit committing to purchasing hydropower and nuclear energy, respectively. But their most recent sustainability reports show that their greenhouse gas pollution is still growing, taking them further away from their climate goals of reaching net zero emissions.

“If [tech companies] wanted to meet their sustainability goals, they could do so,” Kunkel says. “They’re getting a free pass, obviously, from the Trump administration.”

A mushroom casket marks a first for ‘green burials’ in the US

2025-07-20T21:54:20-04:00

Loop Biotech’s “Living Cocoon” is a casket made from mycelium, the root structure of mushrooms.

“I’m probably the only architect who created a final home,” Bob Hendrikx tells The Verge. Tombs and catacombs aside, Hendrikx might be the only one to make a final home using mushrooms.

Hendrikx is the founder and CEO of Loop Biotech, a company that makes caskets out of mycelium, the fibrous root structure of mushrooms. This June, the first burial in North America to use one of Loop Biotech’s caskets took place in Maine.

“He always said he wanted to be buried naked in the woods.”

The mushroom casket gives people one more option to leave the living with a gentler impact, part of a growing array of what are supposed to be more sustainable alternatives to traditional burials. Mycelium has also had a moment in recent years, with other eco-conscious designers making biodegradable packaging, leather, and bricks from the material.

Hendrikx started out trying to make a “living home” from mycelium, a material that can be used to make self-healing structures if the fibers continue to grow. While he was studying architecture at Delft University of Technology, he says someone asked him what would happen if their grandma happened to die in that home.

Bob Hendrikx and Marsya Ancker at a ceremony for her father in Maine. | Photo courtesy of Loop Biotech" data-portal-copyright="Photo courtesy of Loop Biotech" />

“It would be great, because there’s going to be so much positivity for Earth,” he recalls answering and then thinking — “Oh my God, this should be a casket.” The mushroom casket became his graduation project, and Hendrikx started Loop Biotech in the Netherlands in 2021.

The casket, which Loop Biotech calls a “Living Cocoon” and sells for around $4,000, is made entirely of mycelium and can be grown in seven days. It can then biodegrade completely in about 45 days, according to the company. The body inside, however, takes longer. In a typical casket, it could be decades before a body fully decomposes. But since fungi can help break down dead organic matter, that time shortens to two to three years in a Living Cocoon, Hendrikx says.

“I personally hate the idea of a body just lying there in the ground,” says Marsya Ancker, whose father, Mark Ancker, was laid to rest in a Living Cocoon in Maine in June. “I don’t want to lie in the ground, but I’m happy to become part of the soil and feed the plants.” She heard about Loop Biotech in a TED Talk years ago and decided to call up the company the day after she got the call that her dad had passed.

Mark Ancker lays on the Charlie's Angels pinball machine, “of which he was the campus king” at his university, according to his daughter, Marsya Ancker. | Photo courtesy of Marsya Ancker" data-portal-copyright="Photo courtesy of Marsya Ancker" />

“He would have gotten a kick out of it, out of the fact that he was the first [to be buried in a Living Cocoon],” Marsya adds. Her family’s not one to miss an opportunity. Marsya described an iconic photo of her dad sitting on a green Volkswagen bus on the way to Woodstock, looking out over a traffic jam with binoculars, soon after Marsya was born and came home from the hospital. “Don’t be ridiculous,” there’s no sense in wasting both their tickets, Marsya says her mom told her dad.

“He always said he wanted to be buried naked in the woods,” Marsya says. “As a younger person, that horrified me. I’m like, ‘But how will I remember you?’ … This way he gets to be buried naked in the woods.” And she’ll have something there to remember him by; the family planted a memorial garden with some of Mark’s favorite perennials on the land where he was buried. Loop Biotech says its mushroom casket will help enrich the soil below.

Marsya also finds the chemicals used in embalming “gross.” A desire to minimize waste and pollution is another reason some people are turning away from standard caskets or cremation.

Conventional burials in the US use around 4.3 million gallons of embalming fluid, 20 million board feet of hardwood, and 1.6 million tons of reinforced concrete each year, according to the Green Burial Council.

The first Living Cocoon burial in the US (which follows thousands more using Loop Biotech’s mushroom casket in Europe), shows “there’s excitement and energy around green burial,” says Sam Bar, who is part of the board of directors of the Green Burial Council.

@verge
This casket is grown entirely from the root structure of mushrooms. It’s part of a growing green burial movement. Conventional US burials use millions of gallons of embalming fluid and tons of concrete and hardwood each year. Loop Biotech, a Dutch startup, is now bringing a greener alternative to the US for the first time. Read more at TheVerge.com #environment #mushroom #burial #science #tech
♬ original sound – The Verge

A “green” burial doesn’t have to incorporate mushrooms, of course. The goal is primarily to encourage decomposition and use natural materials in a sustainable way, Bar says. That can also be accomplished using other materials that break down more easily, like woven sea grass or bamboo. “Green is a spectrum,” Bar says.

Ever the architect, Hendrikx has also kept comfortable design in mind with his Living Cocoon. Aside from the potential environmental benefits, the mushroom casket is also soft to the touch and rounded, he points out to The Verge. “So instead of having, like, a hard, pointy casket, you now have something that you can actually hug,” Hendrikx says. “Which is really nice for the grieving process.”

How to design an actually good flash flood alert system

2025-07-18T20:38:09-04:00

An aerial view of flash flood damage along the banks of the Guadalupe River on July 11th in Kerrville, Texas. | Photo: Getty Images

Flash floods have wrought more havoc in the US this week, from the Northeast to the Midwest, just weeks after swollen rivers took more than 130 lives across central Texas earlier this month. Frustrations have grown in the aftermath of that catastrophe over why more wasn’t done to warn people in advance.

Local officials face mounting questions over whether they sent too many or sent too few mobile phone alerts to people. Some Texans have accused the state of sending out too many alerts for injured police officers in the months leading up to the floods, which may have led to residents opting out of receiving warnings. And hard-hit Kerr County, where more than 100 people died, lacked sirens along riverbanks to warn people of rising waters.

These are all important questions to answer that can help keep history from repeating itself in another disaster. Failing to translate flood forecasts into timely messages that tell people what they need to do to stay safe can have tragic consequences. In Texas and elsewhere, the solution is more wide-ranging than fixing any single channel of communication. The Verge spoke with experts about what it would take to design an ideal disaster warning system.

The solution is more wide-ranging than fixing any single channel of communication

When you have a matter of hours or maybe even minutes to send a lifesaving message, you need to use every tool at your disposal. That communication needs to start long before the storm rolls in, and involves everyone from forecasters to disaster managers and local officials. Even community members will need to reach out to each other when no one else may be able to get to them.

The last mile

By definition, flash floods are difficult to forecast with specificity or much lead time. But forecasts are only one part of the process. There are more hurdles when it comes to getting those forecasts out to people, an issue experts describe as getting past “the last mile.” Doing so starts with a shift in thinking from “‘what will the weather be’ to ‘what will the weather do,’” explains Olufemi Osidele, CEO of Hydrologic Research Center (HRC), which oversees a global flash flood guidance program. The technical term is “impact-based forecasting,” and the goal is to relay messages that help people understand what actions to take to keep themselves safe.

In the hours leading up to devastating floods in central Texas, the National Weather Service sent out escalating alerts about the growing risk of flash floods. But not everyone received alerts on their phones with safety instructions from Kerr County officials during crucial hours, according to records obtained by NBC News. While meteorologists can say there’s a life-threatening storm approaching, it typically falls to local authorities to determine what guidance to give to specific communities on how and when to evacuate or take shelter.

“Emergency responders need to know what are the appropriate actions to take or what’s needed in the case of a flash flood before an event happens so that they can react quickly, because the time to respond to that event is likely very short,” says Theresa Modrick Hansen, chief operating officer at HRC. “Time is really the critical issue for disaster managers.”

A blank box

Without prior planning, local alerting authorities might be stuck staring at a blank screen when deciding what warning to send to people in the heat of the moment. Many alerting platforms don’t include instructions on how to write that message, according to Jeannette Sutton, an associate professor in the College of Emergency Preparedness, Homeland Security and Cybersecurity at the University at Albany, SUNY. Sutton is also the founder of The Warn Room and consults with local organizations on how to improve their warning systems.

“When you sit down at the keyboard, you have a blank box that you have to fill in with the information that’s going to be useful to the public,” Sutton says. “And when you are in a highly volatile, emotional, chaotic situation, and you all of a sudden have to create [a] message very quickly that is really clear and complete and directed to the right people at the right time, it’s really hard to think of all of that in the moment.”

There aren’t national standards for how a flood alert system should work in the US, so practices vary from place to place. Sutton recommends an end-to-end warning system that connects each step of the process and the people along the way. It includes forecasters and hydrologists who collect data and run it through predictive models to understand the potential impact on communities — identifying which specific populations or infrastructure are most vulnerable. They need to get that information quickly to disaster managers who can then reach people most at risk with safety instructions using channels of communication they’ve thought through in advance.

Ideally, those alerts are tailored to specific locations and give people clear instructions — telling them who should evacuate, when, and where, for instance. A strong message should include five things, according to Sutton: who the message is from, what the hazard is doing, the location and timing of the threat, and what actions to take to protect yourself.

“If you are receiving a warning that’s statewide or county wide, it can be difficult for some people to understand if they should act or evacuate,” says Juliette Murphy, CEO and co-founder of the flood forecasting company FloodMapp. “Or if a warning states that a river will reach 30 feet, that might not mean much to some people if they don’t have a hydrology understanding.”

Murphy’s company is now using its mapping tools to help state and federal agencies find dozens of people still missing since the July 4th floods. FloodMapp hadn’t worked with counties affected by the floods prior to this disaster, but Murphy says she’d like to work with local agencies in the future that want to improve their warning systems.

Kerr County is under scrutiny for lacking flood sirens, even though county commissioners had been talking about the need to upgrade its flood systems — including adding sirens — since at least 2016. The county sits in an area known as “flash flood alley” because of the way the hilly topography of the area heightens flood risk during storms. Sirens in neighboring communities have been credited with saving lives.

“If I were to envision a really good, robust warning system in flash flood alley, I would say that there would be sirens in these very rural, remote areas,” Sutton says.

Sirens can be critical for reaching people outdoors who may not have cell service and are hard to reach. Even so, it’s no silver bullet. The sound doesn’t necessarily reach people indoors who are further from the riverbanks but still in harm’s way. And it doesn’t provide clear instructions on what actions people need to take.

Along with sirens, Sutton says she’d recommend making sure communities are prepared with “call trees” in advance. That means people are physically picking up the phone; each person is responsible for calling three more people, and so on. “It’s the human touch,” Sutton says. In worst-case scenarios, that might include going out to pound on neighbors’ doors. And that human touch can be especially important for reaching someone who might be skeptical of a government agency sending an alert but might trust a friend or fellow church member, for example, or for those who speak a different language than what officials use.

Wireless emergency alerts are also critical; Sutton considers them the most powerful alerting system across the US because it does not require people to opt in to get a message. But there are also warning systems that people can opt in to for alerts, including CodeRed weather warnings. Kerr County used CodeRed to send out warnings to people subscribed to that system, and audio recordings from disaster responders on July 4th have raised more questions about whether those messages were too delayed to keep people out of danger.

In an email to The Verge, a Kerr County spokesperson said the county is committed to “transparency” and a “full review” of the disaster response. State lawmakers start a special session next week and are expected to consider legislation to bolster flood warning systems and emergency communications. One Senate bill would let municipalities gather residents’ contact information to enroll them in text alerts that they could opt out of if they don’t want to receive them.

Disaster fatigue and Swiss cheese

People opting out of notifications has also been a concern — particularly after a deluge of “Blue alerts” sent after a law enforcement officer has been injured or killed. Frustrations have flared up on social media this month over a statewide Blue alert issued for someone suspected of being involved in the “serious injury” of a police officer at an Immigration and Customs Enforcement (ICE) detention facility in Alvarado, Texas. “Texas can’t adequately warn people about deadly floods, but it can immediately let me know that a cop got hurt 250 miles away from me,” one post with more than 20,000 likes on Bluesky says. The FCC has received thousands complaints about the Blue alert system in Texas, CBS News reported in October of last year.

“Alert fatigue” is a concern if it pushes people to ignore warnings or opt out of receiving them altogether. That can be an issue during extreme weather if authorities include Blue alerts and extreme weather warnings in the same “imminent threat” category of wireless emergency alerts. Again, this can vary from locality to locality. “It’s really frustrating when they choose to send a Blue alert through an imminent threat channel,” Sutton says. To stop getting those pings about police officers, someone might opt out of the imminent threat category of wireless emergency alerts — but that means they would also stop getting other alerts in the same channel for weather emergencies.

“This is exactly what we don’t want to have happen, because when you turn it off you’re not going to get the message for that flash flood. So it’s really dangerous,” Sutton says.

“This is exactly what we don’t want to have happen”

Even so, we still don’t have data on who might have missed a lifesaving alert because of frustration with Blue alerts. Nor do we know the extent to which people are just ignoring notifications, or why. The number of public safety alerts sent in Texas has doubled since 2018 for a wide range of warnings, including Blue alerts, Silver alerts for missing elderly adults, Amber alerts for missing children, and more, the Houston Chronicle reports.

And when it comes to warning people about flash floods in particular, experts still stress the need to get warnings to people via every means possible. If someone misses a wireless emergency alert, there should be another way to reach them. There are likely going to be gaps when it comes to any single strategy for alerting people, as well as other complications that can impede the message getting out. (On July 4th, floodwaters rose in the dead of night — making it even harder to notify people as they slept.)

That’s why a “Swiss cheese” approach to warning people can be most effective in overcoming that last mile, Chris Vagasky, a meteorologist and manager of the Wisconsin Environmental Mesonet at the University of Wisconsin-Madison, explains. (And it’s similar to an ideology used to prevent the spread of disease.)

“You know you got slices of Swiss cheese and they’ve got holes in them. Nothing is ever perfect. But if you layer enough pieces of cheese, it reduces the risk because something might go through one hole, but then it gets blocked,” Vagasky says. “We always want people to have multiple ways of receiving warnings.”

Conspiracy theorists are blaming flash floods on cloud seeding — it has to stop

2025-07-10T20:04:12-04:00

As The Verge’s resident disaster writer, I’m tired of this nonsense. So let’s just get into it.

What is cloud seeding?

Cloud seeding is basically an attempt to make precipitation fall from clouds. It targets clouds that have water droplets that are essentially too light to fall. Scientists at MIT learned in the 1940s that if you inject a mineral into the cloud that’s similar to the crystalline structure of ice — typically silver iodide or salt — those small water droplets start to freeze to the mineral. This creates heavier ice particles that can eventually fall down to the ground. These days, researchers can use radar and satellite imagery to identify the right kind of clouds and then fly drones or planes into them to disperse the mineral.

Why are we talking about it now?

Cloud seeding has become a regular scapegoat for devastating flooding

Cloud seeding has become a regular scapegoat for devastating flooding events. After horrific flash floods in central Texas killed at least 120 people over the July 4th weekend, a flurry of social media posts blamed cloud seeding. One startup called Rainmaker has borne the brunt of attacks that have turned into violent threats.

“There have been death threats, both via email and online, and our team has handled that like a bunch of champs,” Rainmaker CEO Augustus Doricko tells The Verge, adding that the company now has security at all of its facilities “out of an abundance of caution.”

This isn’t the first time Rainmaker has faced the repercussions of misinformation about cloud seeding. It cropped up during Hurricane Milton and Hurricane Helene. UCLA climate scientist Daniel Swain held an online “office hour” on YouTube to debunk false claims about cloud seeding following extreme rainfall in Dubai in April 2024.

But the backlash against cloud seeding has been particularly intense in the aftermath of the deadly July 4th flash floods. Doricko attributes that in part to President Donald Trump’s former National Security Advisor Michael Flynn hopping on the bandwagon of lawmakers and right-wing influencers giving credence to the misleading attempts to link cloud seeding to the disaster in Texas. “Anyone who calls this out as a conspiracy theory can go F themselves,” Flynn wrote on X.

Inflaming things further, Rep. Marjorie Taylor Greene (R-GA) announced on July 5th that she would introduce a bill that would make a felony offense out of “the injection, release, or dispersion of chemicals or substances into the atmosphere for the express purpose of altering weather.”

Along a similar vein, conspiracy theories maligning Doppler weather radars as “weather weapons” have heightened in the wake of flash floods in central Texas. Forecasters use the radar system to detect precipitation, and now have to worry about vigilantes vandalizing them at times when people rely on those forecasts to stay safe.

Could cloud seeding have caused flash flooding?

Cloud seeding cannot conjure up a storm.

There isn’t even consensus on how helpful cloud seeding can be to help ease a water shortage. A critique Swain makes of of this strategy is that it can likely only lead to a modest increase in precipitation over a small area at best. So it’s even less likely to have the opposite problem.

Cloud seeding does not add more moisture to the atmosphere than what’s already present, Swain explains in his YouTube office hour. “You’re really just encouraging existing moisture in existing clouds to fall out with a slightly higher efficiency than you did before,” he says. “This is not something that can create storms. It can’t even create clouds.”

Chris Vagasky, a meteorologist and manager of the Wisconsin Environmental Mesonet at the University of Wisconsin-Madison, similarly says in an email to The Verge:

“[Cloud seeding] campaigns usually focus on just a few target clouds and would not have the ability to impact a large area. The amount of energy required to create a complex of thunderstorms and heavy rain is so high that it outweighs the small addition of silver iodide or other seed material.”

Rainmaker has gotten flak from cloud seeding conspiracists for a mission it conducted for clients on July 2nd. It seeded two clouds with about 70 grams of silver iodide (“That’s, like, 10 Skittles’ worth,” Doricko says) for customers seeking to squeeze out more rain for farms below (in the “eastern portions of south-central Texas,” according to Doricko.) Those clouds dissipated within a couple hours, more than a day before the thunderstorms arrived that would inundate the area.

So what did cause the devastating flooding?

A dangerous confluence of heavy rainfall and a hilly landscape funneled water into the Guadalupe River and surrounding areas that quickly turned into deadly rapids on July 4th. Before this catastrophe, the region was already known as “flash flood alley.”

That’s not to say that humans aren’t capable of making disasters like this worse. Climate change intensified the heavy rain that led to deadly flash floods in central Texas, a preliminary study completed by the ClimaMeter project funded by the European Union and the French National Centre for Scientific Research suggests.

Greenhouse gas pollution from fossil fuels is raising global average temperatures. And in a warmer environment, more water can evaporate and then get wrung out in thunderstorms, Vagasky explains. “Climate change goes through and it primes the pump, it primes the atmosphere,” he says. “Climatologists have been saying for years and years that a warmer climate is going to increase the likelihood of these really extreme rainfall events.”

And the more that conspiracy theories distract people from what’s really going on, the harder it is to tackle these problems.

The Bezos-funded climate satellite is lost in space

2025-07-09T14:03:47-04:00

MethaneSat was built to track methane pollution around the world.

A satellite tracking global methane pollution has gone dark, imperiling a mission that garnered enormous support from Jeff Bezos and other big names in tech.

Methane is the primary ingredient of so-called “natural gas” that is even more potent than carbon dioxide when it comes to its ability to heat the planet. The powerful greenhouse gas routinely escapes from oil and gas wells, pipelines, and other fossil fuel infrastructure without anyone seeing or reporting it. MethaneSat was meant to spot such leaks from space in an effort to hold industry accountable for reducing those emissions.

But since June 20th, mission operations haven’t been able to contact MethaneSat. The satellite has lost power and is “likely not recoverable,” according to an update shared today by the nonprofit Environmental Defense Fund that developed MethaneSat.

The satellite has lost power and is “likely not recoverable”

The satellite cost $88 million to build and launch, and the effort received a $100 million grant from the Bezos Earth Fund. It launched in March of last year from a SpaceX Falcon 9 rather than a Blue Origin rocket. The launch marked the first government-funded space mission by New Zealand’s Space Agency, which supported mission operations control and an atmospheric science program.

Before MethaneSat, EDF had to take methane readings on the ground and by aircraft to measure gas leaks. That painstaking work was revelatory; it found that US methane emissions were actually 60 percent higher than the Environmental Protection Agency’s estimates between 2012 and 2018.

Taking readings from space, MethaneSat was supposed to be able to survey an area in about 20 seconds that would have taken an aircraft 2 hours. Orbiting Earth in 95 minutes, it would cover oil and gas fields accounting for more than 80 percent of global production.

Google also partnered with EDF to track methane emissions. With a similar strategy to the way Google Maps identifies sidewalks and street signs in satellite imagery, the company started training AI to spot well pads, pump jacks, storage tanks, and other fossil fuel infrastructure.

EDF says it’s still working to process data MethaneSat has been able to gather since launching, which it hopes can be used to limit methane pollution.

Supreme Court gives Trump the green light to cut down the federal workforce

2025-07-08T18:52:31-04:00

In a significant albeit preliminary win for the president, the Supreme Court is allowing the Trump administration to go ahead with its plans to drastically reshape federal agencies by laying off employees en masse.

The unsigned decision lifts a lower court ruling that briefly stopped the Trump administration from cutting down the federal workforce while a legal challenge worked through the court system. The Supreme Court is now allowing federal agencies to slash jobs even before the legality of Trump’s plan has been decided.

A February executive order directed the heads of federal agencies to “promptly undertake preparations to initiate large-scale reductions in force (RIFs).” A coalition of unions, local governments, and nonprofit organizations soon filed suit, alleging that Trump’s attempt to reorganize the federal government was unlawful.

“Today’s decision has dealt a serious blow to our democracy.”

In May, a US District Court Judge Susan Illston from San Francisco granted a preliminary injunction pausing the mass layoffs while the lawsuit proceeds. “Agencies may not conduct large-scale reorganizations and reductions in force in blatant disregard of Congress’s mandates, and a President may not initiate large-scale executive branch reorganization without partnering with Congress,” the order reads.

The Supreme Court is now lifting that injunction on the basis that the Trump administration is “likely to succeed” in its argument that the executive order is lawful. But it chose not to weigh in on the legality of any individual agency’s RIF plan for now.

“Today’s decision has dealt a serious blow to our democracy and puts services that the American people rely on in grave jeopardy,” the coalition of groups that sued Trump said in a press release today.

Why it’s so hard to warn people about flash floods

2025-07-08T15:06:46-04:00

A photo shows flooding caused by a flash flood at the Guadalupe River in Kerrville, Texas, on July 5th, 2025. | Photo: Getty Images

By definition, flash floods are notoriously difficult to warn people about well in advance. They form rapidly, giving forecasters hours of lead time at best to figure out where they might hit with specificity. We’ve seen this with devastating effect in Texas, where flash floods over Independence Day weekend killed over 100 people — many of them children and families who were in bed when officials issued emergency warnings.

Issuing warnings requires a whole lot of weather and water data. Foreseeing how much rain is likely to fall, and then figuring out the flow of that water on land, are both complicated tasks. Climate change adds another risk factor. Meanwhile, the Trump administration’s cuts to federal climate research and weather forecasting could make an already challenging process more precarious in the future.

What makes a “flash” flood? The National Weather Service (NWS) defines it as flooding that starts within six hours of heavy rainfall or another sudden trigger like a dam or levee break. Storms are usually the culprit. And predicting the amount of water that’s going to fall out of the sky — called quantitative precipitation forecasting — is something that scientists are still working on.

“Getting those very precise measurements at those very precise locations is something that we’re still working on”

The shape of a cloud, where water accumulates in the cloud, and how dry the air is between the cloud and the ground in different locations, are all factors that might influence how much rain hits the ground in a certain location, according to Chris Vagasky, a meteorologist and manager of the Wisconsin Environmental Mesonet at the University of Wisconsin-Madison. The size of raindrops falling, along with wind speed and direction, are also contributing factors. For all these reasons, it’s very hard to say exactly how many inches of rain to expect in a specific location.

“Getting those very precise measurements at those very precise locations is something that we’re still working on, improving that science,” Vagasky says. Progress hinges on more advanced computer modeling and a better understanding of how precipitation forms in clouds.

Flash flood forecasting also involves charting the flow of that water on land, which poses its own set of challenges. The type of terrain it falls on — whether it’s sandy or rocky soil or an urban area with a lot of impervious concrete that prevents the ground from soaking up water, for instance — makes a difference. So does the geography of the land and the size of the watershed into which the rain falls. A watershed or drainage basin is a defined area where rainfall and snowmelt generally flow into the same bodies of water. Where there are canyons and hills, even a relatively small amount of rain over a wide-enough area could lead to flash flooding if all the water is funnelled into the same river. The July 4th flooding took place in the hill country of Texas, in an area around an inactive fault zone called Balcones Escarpment that’s dubbed “flash flood alley” because of the heightened risk here.

Forecasts also have to consider the delay time between when the rainfall is most intense over a particular drainage basin and when peak flooding occurs wherever the water converges. The response time is shorter if the basin is small, steep, or highly urbanized, says Claudio Meier, a water resources engineer and associate professor at the University of Memphis.

“[That] means that from the moment that you’re seeing all this rain falling to the moment you get the big flood, you only get a few tens of minutes to a couple of hours. So that’s very little time to warn people or do anything about it,” Meier says.

A river gauge along the Guadalupe River at Hunt in Texas showed how fast water levels rose to deadly levels. The flow of water climbed from 8 cubic feet per second at 1:10AM to 120,000 cubic feet per second at 4:35AM — just before the gauge failed from the inundation. “Essentially, at 1:10 am the river was a tranquil almost dry riverbed, and by 4:30 am it was a raging flood with more water flowing than the average flow over Niagara Falls,” meteorologist Alan Gerard wrote in his Balanced Weather blog on July 5th.

Prominent scientists have defended the NWS forecasts following claims from some local officials that they didn’t have enough advance notice. Forecasters can warn of excessive rainfall days in advance, but pinpointing precise locations for flash floods requires real-time observations that only allow for hours of notice at most.

The last hurdle is to get these messages in front of people

The NWS issued a flood watch Thursday at 1:18PM to notify people that heavy rainfall could cause flash flooding across portions of eight counties in south-central Texas. A “watch” is an early alert indicating that flooding “is possible.” About 12 hours later, at 1:14AM, the NWS escalated its message, issuing a flash flood warning, which is typically sent out when a flash flood is imminent or already taking place. Thunderstorms were creating “life threatening flash flooding,” the warning said.

The last hurdle is to get these messages in front of people, which Vagasky and other experts call “the last mile.” The catastrophe in Texas arrived in the dead of night when many people were already asleep, making it more difficult to get these warnings out to them.

Getting past that last mile also seems to have gotten tricker recently. X has become a less reliable source of vetted information. And the San Antonio Office of the NWS that played a key role in forecasting was missing a “warning coordination meteorologist” after DOGE cuts to the agency.

“All forecasts and warnings were issued in a timely manner. Additionally, these offices were able to provide decision support services to local partners, including those in the emergency management community,” a spokesperson for the NWS said in an email to The Verge.

Experts The Verge spoke to emphasized how crucial it will be to continue gathering the robust datasets needed to forecast flash floods. The Trump administration’s proposed budget for NWS’ parent agency for the 2026 fiscal year would shutter laboratories and research programs vital to flash flood forecasting, scientists warn. A global Flash Flood Guidance System that helped other countries develop their own warning systems lost funding when DOGE dismantled USAID.

The Trump administration has also dismissed scientists working on a new national assessment of how climate change affects the US. Climate change intensified the heavy rain that led to deadly flash floods in central Texas on July 4th, according to a preliminary study completed by the ClimaMeter project funded by the European Union and the French National Centre for Scientific Research. In a warmer environment, more water can evaporate and then get wrung out in thunderstorms, Vagasky explains.

“It’s absolutely important to rethink how we communicate early warning systems,” says Mireia Ginesta, a research associate at the University of Oxford and one of the authors of the ClimaMeter study. “People should take this more seriously and there absolutely shouldn’t be cuts in funding for research.”