Author name: Kris Guyer

jury-orders-nso-to-pay-$167-million-for-hacking-whatsapp-users

Jury orders NSO to pay $167 million for hacking WhatsApp users

A jury has awarded WhatsApp $167 million in punitive damages in a case the company brought against Israel-based NSO Group for exploiting a software vulnerability that hijacked the phones of thousands of users.

The verdict, reached Tuesday, comes as a major victory not just for Meta-owned WhatsApp but also for privacy- and security-rights advocates who have long criticized the practices of NSO and other exploit sellers. The jury also awarded WhatsApp $444 million in compensatory damages.

Clickless exploit

WhatsApp sued NSO in 2019 for an attack that targeted roughly 1,400 mobile phones belonging to attorneys, journalists, human-rights activists, political dissidents, diplomats, and senior foreign government officials. NSO, which works on behalf of governments and law enforcement authorities in various countries, exploited a critical WhatsApp vulnerability that allowed it to install NSO’s proprietary spyware Pegasus on iOS and Android devices. The clickless exploit worked by placing a call to a target’s app. A target did not have to answer the call to be infected.

“Today’s verdict in WhatsApp’s case is an important step forward for privacy and security as the first victory against the development and use of illegal spyware that threatens the safety and privacy of everyone,” WhatsApp said in a statement. “Today, the jury’s decision to force NSO, a notorious foreign spyware merchant, to pay damages is a critical deterrent to this malicious industry against their illegal acts aimed at American companies and the privacy and security of the people we serve.”

NSO created WhatsApp accounts in 2018 and used them a year later to initiate calls that exploited the critical vulnerability on phones, which, among others, included 100 members of “civil society” from 20 countries, according to an investigation research group Citizen Lab performed on behalf of WhatsApp. The calls passed through WhatsApp servers and injected malicious code into the memory of targeted devices. The targeted phones would then use WhatsApp servers to connect to malicious servers maintained by NSO.

Jury orders NSO to pay $167 million for hacking WhatsApp users Read More »

trump-and-doj-try-to-spring-former-county-clerk-tina-peters-from-prison

Trump and DOJ try to spring former county clerk Tina Peters from prison

President Donald Trump is demanding the release of Tina Peters, a former election official who parroted Trump’s 2020 election conspiracy theories and is serving nine years in prison for compromising the security of election equipment.

In a post on Truth Social last night, Trump wrote that “Radical Left Colorado Attorney General Phil Weiser ignores Illegals committing Violent Crimes like Rape and Murder in his State and, instead, jailed Tina Peters, a 69-year-old Gold Star mother who worked to expose and document Democrat Election Fraud. Tina is an innocent Political Prisoner being horribly and unjustly punished in the form of Cruel and Unusual Punishment.”

Trump said he is “directing the Department of Justice to take all necessary action to help secure the release of this ‘hostage’ being held in a Colorado prison by the Democrats, for political reasons.”

The former Mesa County clerk was indicted in March 2022 on charges related to the leak of voting-system BIOS passwords and other confidential information. Peters was convicted in August 2024 and later sentenced in a Colorado state court.

“Your lies are well-documented and these convictions are serious,” 21st Judicial District Judge Matthew Barrett told Peters at her October 2024 sentencing. “I am convinced you would do it all over again. You are as defiant a defendant as this court has ever seen.”

DOJ reviews case for “abuse” of process

After Peters’ August 2024 conviction, Colorado Secretary of State Jena Griswold said that “Tina Peters willfully compromised her own election equipment trying to prove Trump’s big lie.”

Peters appealed her conviction in a Colorado appeals court and separately sought relief in US District Court for the District of Colorado. She asked the federal court to order her release on bond while the state court system handles her appeal and said her health has deteriorated while being incarcerated.

Trump’s Justice Department submitted a filing on Peters’ behalf in March, saying the US has concerns about “the exceptionally lengthy sentence imposed relative to the conduct at issue, the First Amendment implications of the trial court’s October 2024 assertions relating to Ms. Peters, and whether Colorado’s denial of bail pending appeal was arbitrary or unreasonable under the Eighth and Fourteenth Amendments.”

Trump and DOJ try to spring former county clerk Tina Peters from prison Read More »

2025-alfa-romeo-tonale-turbo-review:-italian-charm-that-cuts-both-ways

2025 Alfa Romeo Tonale Turbo review: Italian charm that cuts both ways

While it’s nice to see that this feature is standard equipment, the ACC system can be a bit overeager to close the gap between you and the car in front of you, and it has a bad habit of braking later than it should. This resulted in several panic stops where the adaptive cruise control’s behavior triggered the forward collision warning system despite the fact that I was plodding along at just 15 mph (24 km/h) stop-and-go traffic.

In a canyon with some paddles

Out in the canyons, I switched to the Dynamic drive mode, which sharpens the Tonale’s reflexes and adds more urgency to the proceedings. The dampers’ added stiffness in this mode cleaned up the crossover’s body control to a tangible degree, but the transmission’s ongoing search for more efficient gears was only alleviated by switching to manual mode and taking over control of the gearbox with the paddles. While the overall tuning is a bit softer than some enthusiasts will prefer, the Tonale’s performance is buoyed by a gutsy powerplant and the confident stopping power delivered by the four-piston Brembo brakes equipped up front.

With a starting price of $36,495 ($48,130 as-tested, with destination fee), the Tonale 2.0 L Turbo is roughly $10,000 cheaper than its hybrid counterpart, and it also undercuts other European premium compact crossovers like the Mercedes-Benz GLA and BMW X1 by thousands. The Tonale’s extroverted character is also a nice change of pace in a segment filled with anonymity, and given the negligible compromise in straight-line performance, the lower curb weight, and the significant cost savings, I’d choose the 2.0 L Turbo over the PHEV model without hesitation.

Telephone dial wheels ftw. Alfa Romeo

The Tonale 2.0 L Turbo’s biggest rival is arguably the Dodge Hornet GT, which offers a similar driving experience but starts at a base price that’s roughly $6,000 lower. The premium you’ll pay for the Tonale largely comes down to its Italian aesthetic and the sense of occasion that the Alfa Romeo name imparts. Those attributes may seem trivial at first glance, but one should never underestimate the value of style.

2025 Alfa Romeo Tonale Turbo review: Italian charm that cuts both ways Read More »

a-doge-recruiter-is-staffing-a-project-to-deploy-ai-agents-across-the-us-government

A DOGE recruiter is staffing a project to deploy AI agents across the US government


“does it still require Kremlin oversight?

A startup founder said that AI agents could do the work of tens of thousands of government employees.

An aide sets up a poster depicting the logo for the DOGE Caucus before a news conference in Washington, DC. Credit: Andrew Harnik/Getty Images

A young entrepreneur who was among the earliest known recruiters for Elon Musk’s so-called Department of Government Efficiency (DOGE) has a new, related gig—and he’s hiring. Anthony Jancso, cofounder of AcclerateX, a government tech startup, is looking for technologists to work on a project that aims to have artificial intelligence perform tasks that are currently the responsibility of tens of thousands of federal workers.

Jancso, a former Palantir employee, wrote in a Slack with about 2000 Palantir alumni in it that he’s hiring for a “DOGE orthogonal project to design benchmarks and deploy AI agents across live workflows in federal agencies,” according to an April 21 post reviewed by WIRED. Agents are programs that can perform work autonomously.

We’ve identified over 300 roles with almost full-process standardization, freeing up at least 70k FTEs for higher-impact work over the next year,” he continued, essentially claiming that tens of thousands of federal employees could see many aspects of their job automated and replaced by these AI agents. Workers for the project, he wrote, would be based on site in Washington, DC, and would not require a security clearance; it isn’t clear for whom they would work. Palantir did not respond to requests for comment.

The post was not well received. Eight people reacted with clown face emojis, three reacted with a custom emoji of a man licking a boot, two reacted with custom emoji of Joaquin Phoenix giving a thumbs down in the movie Gladiator, and three reacted with a custom emoji with the word “Fascist.” Three responded with a heart emoji.

“DOGE does not seem interested in finding ‘higher impact work’ for federal employees,” one person said in a comment that received 11 heart reactions. “You’re complicit in firing 70k federal employees and replacing them with shitty autocorrect.”

“Tbf we’re all going to be replaced with shitty autocorrect (written by chatgpt),” another person commented, which received one “+1” reaction.

“How ‘DOGE orthogonal’ is it? Like, does it still require Kremlin oversight?” another person said in a comment that received five reactions with a fire emoji. “Or do they just use your credentials to log in later?”

AccelerateX was originally called AccelerateSF, which VentureBeat reported in 2023 had received support from OpenAI and Anthropic. In its earliest incarnation, AccelerateSF hosted a hackathon for AI developers aimed at using the technology to solve San Francisco’s social problems. According to a 2023 Mission Local story, for instance, Jancso proposed that using large language models to help businesses fill out permit forms to streamline the construction paperwork process might help drive down housing prices. (OpenAI did not respond to a request for comment. Anthropic spokesperson Danielle Ghiglieri tells WIRED that the company “never invested in AccelerateX/SF,” but did sponsor a hackathon AccelerateSF hosted in 2023 by providing free access to its API usage at a time when its Claude API “was still in beta.”)

In 2024, the mission pivoted, with the venture becoming known as AccelerateX. In a post on X announcing the change, the company posted, “Outdated tech is dragging down the US Government. Legacy vendors sell broken systems at increasingly steep prices. This hurts every American citizen.” AccelerateX did not respond to a request for comment.

According to sources with direct knowledge, Jancso disclosed that AccelerateX had signed a partnership agreement with Palantir in 2024. According to the LinkedIn of someone described as one of AccelerateX’s cofounders, Rachel Yee, the company looks to have received funding from OpenAI’s Converge 2 Accelerator. Another of AccelerateSF’s cofounders, Kay Sorin, now works for OpenAI, having joined the company several months after that hackathon. Sorin and Yee did not respond to requests for comment.

Jancso’s cofounder, Jordan Wick, a former Waymo engineer, has been an active member of DOGE, appearing at several agencies over the past few months, including the Consumer Financial Protection Bureau, National Labor Relations Board, the Department of Labor, and the Department of Education. In 2023, Jancso attended a hackathon hosted by ScaleAI; WIRED found that another DOGE member, Ethan Shaotran, also attended the same hackathon.

Since its creation in the first days of the second Trump administration, DOGE has pushed the use of AI across agencies, even as it has sought to cut tens of thousands of federal jobs. At the Department of Veterans Affairs, a DOGE associate suggested using AI to write code for the agency’s website; at the General Services Administration, DOGE has rolled out the GSAi chatbot; the group has sought to automate the process of firing government employees with a tool called AutoRIF; and a DOGE operative at the Department of Housing and Urban Development is using AI tools to examine and propose changes to regulations. But experts say that deploying AI agents to do the work of 70,000 people would be tricky if not impossible.

A federal employee with knowledge of government contracting, who spoke to WIRED on the condition of anonymity because they were not authorized to speak to the press, says, “A lot of agencies have procedures that can differ widely based on their own rules and regulations, and so deploying AI agents across agencies at scale would likely be very difficult.”

Oren Etzioni, cofounder of the AI startup Vercept, says that while AI agents can be good at doing some things—like using an internet browser to conduct research—their outputs can still vary widely and be highly unreliable. For instance, customer service AI agents have invented nonexistent policies when trying to address user concerns. Even research, he says, requires a human to actually make sure what the AI is spitting out is correct.

“We want our government to be something that we can rely on, as opposed to something that is on the absolute bleeding edge,” says Etzioni. “We don’t need it to be bureaucratic and slow, but if corporations haven’t adopted this yet, is the government really where we want to be experimenting with the cutting edge AI?”

Etzioni says that AI agents are also not great 1-1 fits for job replacements. Rather, AI is able to do certain tasks or make others more efficient, but the idea that the technology could do the jobs of 70,000 employees would not be possible. “Unless you’re using funny math,” he says, “no way.”

Jancso, first identified by WIRED in February, was one of the earliest recruiters for DOGE in the months before Donald Trump was inaugurated. In December, Jancso, who sources told WIRED said he had been recruited by Steve Davis, president of the Musk-founded Boring Company and a current member of DOGE, used the Palantir alumni group to recruit DOGE members. On December 2nd, 2024, he wrote, “I’m helping Elon’s team find tech talent for the Department of Government Efficiency (DOGE) in the new admin. This is a historic opportunity to build an efficient government, and to cut the federal budget by 1/3. If you’re interested in playing a role in this mission, please reach out in the next few days.”

According to one source at SpaceX, who asked to remain anonymous as they are not authorized to speak to the press, Jancso appeared to be one of the DOGE members who worked out of the company’s DC office in the days before inauguration along with several other people who would constitute some of DOGE’s earliest members. SpaceX did not respond to a request for comment.

Palantir was cofounded by Peter Thiel, a billionaire and longtime Trump supporter with close ties to Musk. Palantir, which provides data analytics tools to several government agencies including the Department of Defense and the Department of Homeland Security, has received billions of dollars in government contracts. During the second Trump administration, the company has been involved in helping to build a “mega API” to connect data from the Internal Revenue Service to other government agencies, and is working with Immigration and Customs Enforcement to create a massive surveillance platform to identify immigrants to target for deportation.

This story originally appeared at WIRED.com.

Photo of WIRED

Wired.com is your essential daily guide to what’s next, delivering the most original and complete take you’ll find anywhere on innovation’s impact on technology, science, business and culture.

A DOGE recruiter is staffing a project to deploy AI agents across the US government Read More »

claude’s-ai-research-mode-now-runs-for-up-to-45-minutes-before-delivering-reports

Claude’s AI research mode now runs for up to 45 minutes before delivering reports

Still, the report contained a direct quote statement from William Higinbotham that appears to combine quotes from two sources not cited in the source list. (One must always be careful with confabulated quotes in AI because even outside of this Research mode, Claude 3.7 Sonnet tends to invent plausible ones to fit a narrative.) We recently covered a study that showed AI search services confabulate sources frequently, and in this case, it appears that the sources Claude Research surfaced, while real, did not always match what is stated in the report.

There’s always room for interpretation and variation in detail, of course, but overall, Claude Research did a relatively good job crafting a report on this particular topic. Still, you’d want to dig more deeply into each source and confirm everything if you used it as the basis for serious research. You can read the full Claude-generated result as this text file, saved in markdown format. Sadly, the markdown version does not include the source URLS found in the Claude web interface.

Integrations feature

Anthropic also announced Thursday that it has broadened Claude’s data access capabilities. In addition to web search and Google Workspace integration, Claude can now search any connected application through the company’s new “Integrations” feature. The feature reminds us somewhat of OpenAI’s ChatGPT Plugins feature from March 2023 that aimed for similar connections, although the two features work differently under the hood.

These Integrations allow Claude to work with remote Model Context Protocol (MCP) servers across web and desktop applications. The MCP standard, which Anthropic introduced last November and we covered in April, connects AI applications to external tools and data sources.

At launch, Claude supports Integrations with 10 services, including Atlassian’s Jira and Confluence, Zapier, Cloudflare, Intercom, Asana, Square, Sentry, PayPal, Linear, and Plaid. The company plans to add more partners like Stripe and GitLab in the future.

Each integration aims to expand Claude’s functionality in specific ways. The Zapier integration, for instance, reportedly connects thousands of apps through pre-built automation sequences, allowing Claude to automatically pull sales data from HubSpot or prepare meeting briefs based on calendar entries. With Atlassian’s tools, Anthropic says that Claude can collaborate on product development, manage tasks, and create multiple Confluence pages and Jira work items simultaneously.

Anthropic has made its advanced Research and Integrations features available in beta for users on Max, Team, and Enterprise plans, with Pro plan access coming soon. The company has also expanded its web search feature (introduced in March) to all Claude users on paid plans globally.

Claude’s AI research mode now runs for up to 45 minutes before delivering reports Read More »

some-flies-go-insomniac-to-ward-off-parasites

Some flies go insomniac to ward off parasites

Those genes associated with metabolism were upregulated, meaning they showed an increase in activity. An observed loss of body fat and protein reserves was evidently a trade-off for resistance to mites. This suggests there was increased lipolysis, or the breakdown of fats, and proteolysis, the breakdown of proteins, in resistant lines of flies.

Parasite paranoia

The depletion of nutrients could make fruit flies less likely to survive even without mites feeding off them, but their tenaciousness when it comes to staying up through the night suggests that being parasitized by mites is still the greater risk. Because mite-resistant flies did not sleep, their oxygen consumption and activity also increased during the night to levels no different from those of control group flies during the day.

Keeping mites away involves moving around so the fly can buzz off if mites crawl too close. Knowing this, Benoit wanted to see what would happen if the resistant flies’ movement was restricted. It was doom. When the flies were restrained, the mite-resistant flies were as susceptible to mites as the controls. Activity alone was important for resisting mites.

Since mites are ectoparasites, or external parasites (as opposed to internal parasites like tapeworms), potential hosts like flies can benefit from hypervigilance. Sleep is typically beneficial to a host invaded by an internal parasite because it increases the immune response. Unfortunately for the flies, sleeping would only make them an easy meal for mites. Keeping both stereoscopic eyes out for an external parasite means there is no time left for sleep.

“The pattern of reduced sleep likely allows the flies to be more responsive during encounters with mites during the night,” the researchers said in their study, which was recently published in Biological Timing and Sleep. “There could be differences in sleep occurring during the day, but these differences may be less important as D. melanogaster sleeps much less during the day.”

Fruit flies aren’t the only creatures with sleep patterns that parasites disrupt. Evidence of shifts in sleep and rest in birds and bats has been shown to happen when there is a risk of parasitism after dark. For the flies, exhaustion has the upside of better fertility if they manage to avoid bites, so a mate must be worth all those sleepless nights.

Biological Timing and Sleep, 2025.  DOI: 10.1038/s44323-025-00031-7

Some flies go insomniac to ward off parasites Read More »

we-finally-know-a-little-more-about-amazon’s-super-secret-satellites

We finally know a little more about Amazon’s super-secret satellites

“Elon thinks we can do the job with cheaper and simpler satellites, sooner,” a source told Reuters at the time of Badyal’s dismissal. Earlier in 2018, SpaceX launched a pair of prototype cube-shaped Internet satellites for demonstrations in orbit. Then, less than a year after firing Badyal, Musk’s company launched the first full stack of Starlink satellites, debuting the now-standard flat-panel design.

In a post Friday on LinkedIn, Badyal wrote the Kuiper satellites have had “an entirely nominal start” to their mission. “We’re just over 72 hours into our first full-scale Kuiper mission, and the adrenaline is still high.”

The Starlink and Kuiper constellations use laser inter-satellite links to relay Internet signals from node-to-node across their networks. Starlink broadcasts consumer broadband in Ku-band frequencies, while Kuiper will use Ka-band.

Ultimately, SpaceX’s simplified Starlink deployment architecture has fewer parts and eliminates the need for a carrier structure. This allows SpaceX to devote a higher share of the rocket’s mass and volume capacity to the Starlink satellites themselves, replacing dead weight with revenue-earning capability. The dispenser architecture used by Amazon is a more conventional design, and gives satellite engineers more flexibility in designing their spacecraft. It also allows satellites to spread out faster in orbit.

Others involved in the broadband megaconstellation rush have copied SpaceX’s architecture.

China’s Qianfan, or Thousand Sails, satellites have a “standardized and modular” flat-panel design that “meets the needs of stacking multiple satellites with one rocket,” according to the company managing the constellation. While Chinese officials haven’t released any photos of the satellites, which could eventually number more than 14,000, this sounds a lot like the design of SpaceX’s Starlink satellites.

Another piece of information released by United Launch Alliance helps us arrive at an estimate of the mass of each Kuiper satellite. The collection of 27 satellites that launched earlier this week added up to be the heaviest payload ever flown on ULA’s Atlas V rocket. ULA said the total payload the Atlas V delivered to orbit was about 34,000 pounds, equivalent to roughly 15.4 metric tons.

It wasn’t clear whether this number accounted for the satellite dispenser, which likely weighed somewhere in the range of 1,000 to 2,000 pounds at launch. This would put the mass of each Kuiper satellite somewhere between 1,185 and 1,259 pounds (537 and 571 kilograms).

This is not far off the estimated mass of SpaceX’s most recent iteration of Starlink satellites, a version known as V2 Mini Optimized. SpaceX’s Falcon 9 rocket has launched up to 28 of these flat-packed satellites on a single launch.

We finally know a little more about Amazon’s super-secret satellites Read More »

judge-on-meta’s-ai-training:-“i-just-don’t-understand-how-that-can-be-fair-use”

Judge on Meta’s AI training: “I just don’t understand how that can be fair use”


Judge downplayed Meta’s “messed up” torrenting in lawsuit over AI training.

A judge who may be the first to rule on whether AI training data is fair use appeared skeptical Thursday at a hearing where Meta faced off with book authors over the social media company’s alleged copyright infringement.

Meta, like most AI companies, holds that training must be deemed fair use, or else the entire AI industry could face immense setbacks, wasting precious time negotiating data contracts while falling behind global rivals. Meta urged the court to rule that AI training is a transformative use that only references books to create an entirely new work that doesn’t replicate authors’ ideas or replace books in their markets.

At the hearing that followed after both sides requested summary judgment, however, Judge Vince Chhabria pushed back on Meta attorneys arguing that the company’s Llama AI models posed no threat to authors in their markets, Reuters reported.

“You have companies using copyright-protected material to create a product that is capable of producing an infinite number of competing products,” Chhabria said. “You are dramatically changing, you might even say obliterating, the market for that person’s work, and you’re saying that you don’t even have to pay a license to that person.”

Declaring, “I just don’t understand how that can be fair use,” the shrewd judge apparently stoked little response from Meta’s attorney, Kannon Shanmugam, apart from a suggestion that any alleged threat to authors’ livelihoods was “just speculation,” Wired reported.

Authors may need to sharpen their case, which Chhabria warned could be “taken away by fair use” if none of the authors suing, including Sarah Silverman, Ta-Nehisi Coates, and Richard Kadrey, can show “that the market for their actual copyrighted work is going to be dramatically affected.”

Determined to probe this key question, Chhabria pushed authors’ attorney, David Boies, to point to specific evidence of market harms that seemed noticeably missing from the record.

“It seems like you’re asking me to speculate that the market for Sarah Silverman’s memoir will be affected by the billions of things that Llama will ultimately be capable of producing,” Chhabria said. “And it’s just not obvious to me that that’s the case.”

But if authors can prove fears of market harms are real, Meta might struggle to win over Chhabria, and that could set a precedent impacting copyright cases challenging AI training on other kinds of content.

The judge repeatedly appeared to be sympathetic to authors, suggesting that Meta’s AI training may be a “highly unusual case” where even though “the copying is for a highly transformative purpose, the copying has the high likelihood of leading to the flooding of the markets for the copyrighted works.”

And when Shanmugam argued that copyright law doesn’t shield authors from “protection from competition in the marketplace of ideas,” Chhabria resisted the framing that authors weren’t potentially being robbed, Reuters reported.

“But if I’m going to steal things from the marketplace of ideas in order to develop my own ideas, that’s copyright infringement, right?” Chhabria responded.

Wired noted that he asked Meta’s lawyers, “What about the next Taylor Swift?” If AI made it easy to knock off a young singer’s sound, how could she ever compete if AI produced “a billion pop songs” in her style?

In a statement, Meta’s spokesperson reiterated the company’s defense that AI training is fair use.

“Meta has developed transformational open source AI models that are powering incredible innovation, productivity, and creativity for individuals and companies,” Meta’s spokesperson said. “Fair use of copyrighted materials is vital to this. We disagree with Plaintiffs’ assertions, and the full record tells a different story. We will continue to vigorously defend ourselves and to protect the development of GenAI for the benefit of all.”

Meta’s torrenting seems “messed up”

Some have pondered why Chhabria appeared so focused on market harms, instead of hammering Meta for admittedly illegally pirating books that it used for its AI training, which seems to be obvious copyright infringement. According to Wired, “Chhabria spoke emphatically about his belief that the big question is whether Meta’s AI tools will hurt book sales and otherwise cause the authors to lose money,” not whether Meta’s torrenting of books was illegal.

The torrenting “seems kind of messed up,” Chhabria said, but “the question, as the courts tell us over and over again, is not whether something is messed up but whether it’s copyright infringement.”

It’s possible that Chhabria dodged the question for procedural reasons. In a court filing, Meta argued that authors had moved for summary judgment on Meta’s alleged copying of their works, not on “unsubstantiated allegations that Meta distributed Plaintiffs’ works via torrent.”

In the court filing, Meta alleged that even if Chhabria agreed that the authors’ request for “summary judgment is warranted on the basis of Meta’s distribution, as well as Meta’s copying,” that the authors “lack evidence to show that Meta distributed any of their works.”

According to Meta, authors abandoned any claims that Meta’s seeding of the torrented files served to distribute works, leaving only claims about Meta’s leeching. Meta argued that the authors “admittedly lack evidence that Meta ever uploaded any of their works, or any identifiable part of those works, during the so-called ‘leeching’ phase,” relying instead on expert estimates based on how torrenting works.

It’s also possible that for Chhabria, the torrenting question seemed like an unnecessary distraction. Former Meta attorney Mark Lumley, who quit the case earlier this year, told Vanity Fair that the torrenting was “one of those things that sounds bad but actually shouldn’t matter at all in the law. Fair use is always about uses the plaintiff doesn’t approve of; that’s why there is a lawsuit.”

Lumley suggested that court cases mulling fair use at this current moment should focus on the outputs, rather than the training. Citing the ruling in a case where Google Books scanning books to share excerpts was deemed fair use, Lumley argued that “all search engines crawl the full Internet, including plenty of pirated content,” so there’s seemingly no reason to stop AI crawling.

But the Copyright Alliance, a nonprofit, non-partisan group supporting the authors in the case, in a court filing alleged that Meta, in its bid to get AI products viewed as transformative, is aiming to do the opposite. “When describing the purpose of generative AI,” Meta allegedly strives to convince the court to “isolate the ‘training’ process and ignore the output of generative AI,” because that’s seemingly the only way that Meta can convince the court that AI outputs serve “a manifestly different purpose from Plaintiffs’ books,” the Copyright Alliance argued.

“Meta’s motion ignores what comes after the initial ‘training’—most notably the generation of output that serves the same purpose of the ingested works,” the Copyright Alliance argued. And the torrenting question should matter, the group argued, because unlike in Google Books, Meta’s AI models are apparently training on pirated works, not “legitimate copies of books.”

Chhabria will not be making a snap decision in the case, planning to take his time and likely stressing not just Meta, but every AI company defending training as fair use the longer he delays. Understanding that the entire AI industry potentially has a stake in the ruling, Chhabria apparently sought to relieve some tension at the end of the hearing with a joke, Wired reported.

 “I will issue a ruling later today,” Chhabria said. “Just kidding! I will take a lot longer to think about it.”

Photo of Ashley Belanger

Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.

Judge on Meta’s AI training: “I just don’t understand how that can be fair use” Read More »

the-2025-aston-martin-vantage:-achingly-beautiful-and-thrilling-to-drive

The 2025 Aston Martin Vantage: Achingly beautiful and thrilling to drive

The new infotainment human-machine interface was a little confusing at first; pairing my phone took about 10 minutes but worked out in the end, and wireless Apple CarPlay behaved itself throughout the week. When starting the car there was sometimes a lengthy wait for the infotainment to boot up, as if the hardware wasn’t quite powerful enough for the software stack on top.

Don’t get it wet?

Those are very minor complaints compared to what happened when it rained really hard, though. I had been planning to drive the Vantage to dinner, but upon getting in the car and turning it on, I realized it had other ideas. A number of yellow glyphs remained brightly illuminated on the digital main display—including traction control. A series of notifications informed me of what else wasn’t working, including the ability to switch the car into Wet mode.

Inclement weather. Jonathan Gitlin

With the numbers “$264,300” and “656 hp” at the forefront of my mind, the DC metro sufficed for that trip.

Later that evening, the air still damp but no longer raining, it was if it had all been a dream. A push of the start button and everything lit up and then went dim as normal. The infotainment even booted in an acceptable time. My guess is that there was an improperly sealed connector somewhere, and in torrential rain, water got where it shouldn’t. Regardless, the car let me down when it was supposed to take me somewhere.

As a former resident of the UK, it saddens me to no end that the stereotype about British cars and their electronics is as true today as ever, but the fact remains that every British car I drive, from Minis to McLarens, manages to have some kind of digital or electrical foible that should have been ironed out. Do better, Britain!

With a base price of $191,000, the Vantage is competing with a number of other sports cars, but its biggest rival has to be the Porsche 911 Turbo. It’s a much more dramatic car than the Porsche, in all the meanings of that word, both good and bad.

Aston Martin Vantage seen from behind

Credit: Jonathan Gitlin

Sure looks good though.

The 2025 Aston Martin Vantage: Achingly beautiful and thrilling to drive Read More »

raspberry-pi-cuts-product-returns-by-50%-by-changing-up-its-pin-soldering

Raspberry Pi cuts product returns by 50% by changing up its pin soldering

Getting the hang of through-hole soldering is tricky for those of us tinkering at home with our irons, spools, flux, and, sometimes, braids. It’s almost reassuring, then, to learn that through-hole soldering was also a pain for a firm that has made more than 60 million products with it.

Raspberry Pi boards have a combination of surface-mount devices (SMDs) and through-hole bits. SMDs allow for far more tiny chips, resistors, and other bits to be attached to boards by their tiny pins, flat contacts, solder balls, or other connections. For those things that are bigger, or subject to rough forces like clumsy human hands, through-hole soldering is still required, with leads poked through a connective hole and solder applied to connect and join them securely.

The Raspberry Pi board has a 40-pin GPIO header on it that needs through-hole soldering, along with bits like the Ethernet and USB ports. These require robust solder joints, which can’t be done the same way as with SMT (surface-mount technology) tools. “In the early days of Raspberry Pi, these parts were inserted by hand, and later by robotic placement,” writes Roger Thornton, director of applications for Raspberry Pi, in a blog post. The boards then had to go through a follow-up wave soldering step.

Now Pi boards have their tiny bits and bigger pieces soldered at the same time through an intrusive reflow soldering process undertaken with Raspberry Pi’s UK manufacturing partner, Sony. After adjusting component placement, the solder stencil, and the connectors, the board makers could then place and secure all their components in the same stage.

Raspberry Pi cuts product returns by 50% by changing up its pin soldering Read More »

rfk-jr.-rejects-cornerstone-of-health-science:-germ-theory

RFK Jr. rejects cornerstone of health science: Germ theory


In his 2021 book vilifying Anthony Fauci, RFK Jr. lays out support for an alternate theory.

Health and Human Services Secretary Robert F. Kennedy Jr. speaks at a news conference on removing synthetic dyes from America’s food supply, at the Health and Human Services Headquarters in Washington, DC on April 22, 2025. Credit: Getty | Nathan Posner

With the rise of Robert F. Kennedy Jr., brain worms have gotten a bad rap.

A year ago, the long-time anti-vaccine advocate and current US health secretary famously told The New York Times that a parasitic worm “got into my brain and ate a portion of it and then died.” The startling revelation is now frequently referenced whenever Kennedy says something outlandish, false, or offensive—which is often. For those who have followed his anti-vaccine advocacy, it’s frightfully clear that, worm-infested or not, Kennedy’s brain is marinated in wild conspiracy theories and dangerous misinformation.

While it’s certainly possible that worm remnants could impair brain function, it remains unknown if the worm is to blame for Kennedy’s cognitive oddities. For one thing, he was also diagnosed with mercury poisoning, which can cause brain damage, too. As prominent infectious disease expert Anthony Fauci said last June in a conversation with political analyst David Axelrod: “I don’t know what’s going on in [Kennedy’s] head, but it’s not good.”

The trouble is that now that Kennedy is the country’s top health official, his warped ideas are contributing to the rise of a dystopian reality. Federal health agencies are spiraling into chaos, and critical public health services for Americans have been brutally slashed, dismantled, or knee-capped—from infectious disease responses, the lead poisoning team, and Meals on Wheels to maternal health programs and anti-smoking initiatives, just to name a few. The health of the nation is at stake; the struggle to understand what goes on in Kennedy’s head is vital.

While we may never have definitive answers on his cognitive situation, one thing is plain: Kennedy’s thoughts and actions make a lot more sense when you realize he doesn’t believe in a foundational scientific principle: germ theory.

Dueling theories

Germ theory is, of course, the 19th-century proven idea that microscopic germs—pathogenic viruses, bacteria, parasites, and fungi—cause disease. It supplanted the leading explanation of disease at the time, the miasma theory, which suggests that diseases are caused by miasma, that is, noxious mists and vapors, or simply bad air arising from decaying matter, such as corpses, sewage, or rotting vegetables. While the miasma theory was abandoned, it is credited with spurring improvements in sanitation and hygiene—which, of course, improve health because they halt the spread of germs, the cause of diseases.

Germ theory also knocks back a lesser-known idea called the terrain theory, which we’ve covered before. This is a somewhat ill-defined theory that generally suggests diseases stem from imbalances in the internal “terrain” of the body, such as malnutrition or the presence of toxic substances. The theory is linked to ideas by French scientist Antoine Béchamp and French physiologist Claude Bernard.

Béchamp, considered a bitter crank and rival to famed French microbiologist Louis Pasteur, is perhaps best known for wrongly suggesting the basic unit of organisms is not the cell, but nonexistent microanatomical elements he called “microzyma.” While the idea was largely ignored by the scientific community, Béchamp suggested that disruptions to microzyma are a predisposition to disease, as is the state of the body’s “terrain.” French physiologist Claude Bernard, meanwhile, came up with an idea of balance or stability of the body’s internal environment (milieu intérieur), which was a precursor to the concept of homeostasis. Ideas from the two figures came together to create an ideology that has been enthusiastically adopted by modern-day germ theory denialists, including Kennedy.

It’s important to note here that our understanding of Kennedy’s disbelief in germ theory isn’t based on speculation or deduction; it’s based on Kennedy’s own words. He wrote an entire section on it in his 2021 book vilifying Fauci, titled The Real Anthony Fauci. The section is titled “Miasma vs. Germ Theory,” in the chapter “The White Man’s Burden.”

But, we did reach out to Health and Human Services to ask how Kennedy’s disbelief in germ theory influences his policy decisions. HHS did not respond.

Kennedy’s beliefs

In the chapter, Kennedy promotes the “miasma theory” but gets the definition completely wrong. Instead of actual miasma theory, he describes something more like terrain theory. He writes: “‘Miasma theory’ emphasizes preventing disease by fortifying the immune system through nutrition and by reducing exposures to environmental toxins and stresses.”

Kennedy contrasts his erroneous take on miasma theory with germ theory, which he derides as a tool of the pharmaceutical industry and pushy scientists to justify selling modern medicines. The abandonment of miasma theory, Kennedy bemoans, realigned health and medical institutions to “the pharmaceutical paradigm that emphasized targeting particular germs with specific drugs rather than fortifying the immune system through healthy living, clean water, and good nutrition.”

According to Kennedy, germ theory gained popularity, not because of the undisputed evidence supporting it, but by “mimicking the traditional explanation for disease—demon possession—giving it a leg up over miasma.”

To this day, Kennedy writes, a “$1 trillion pharmaceutical industry pushing patented pills, powders, pricks, potions, and poisons, and the powerful professions of virology and vaccinology led by ‘Little Napoleon’ himself, Anthony Fauci, fortify the century-old predominance of germ theory.”

In all, the chapter provides a clear explanation of why Kennedy relentlessly attacks evidence-based medicines; vilifies the pharmaceutical industry; suggests HIV doesn’t cause AIDS and antidepressants are behind mass shootings; believes that vaccines are harmful, not protective; claims 5G wireless networks cause cancer; suggests chemicals in water are changing children’s gender identities; and is quick to promote supplements to prevent and treat diseases, such as recently recommending vitamin A for measles and falsely claiming children who die from the viral infection are malnourished.

A religious conviction

For some experts, the chapter was like a light bulb going on. “I thought ‘it now all makes sense’… I mean, it all adds up,” Paul Offit, pediatrician and infectious disease expert at Children’s Hospital of Philadelphia, told Ars Technica. It’s still astonishing, though, he added. “It’s so unbelievable, because you can’t imagine that someone who’s the head of Health and Human Services doesn’t believe that specific viruses or bacteria cause specific diseases, and that the prevention or treatment of them is lifesaving.”

Offit has a dark history with Kennedy. Around 20 years ago, Kennedy called Offit out of the blue to talk with him about vaccine safety. Offit knows a lot about it—he’s not only an expert on vaccines, he’s the co-inventor of one. The vaccine he co-developed, RotaTeq, protects against rotaviruses, which cause deadly diarrheal disease in young children and killed an estimated 500,000 people worldwide each year before vaccines were available. RotaTeq has been proven safe and effective and is credited with saving tens of thousands of lives around the world each year.

Kennedy and Offit spent about an hour talking, mostly about thimerosal, an ethylmercury-containing preservative that was once used in childhood vaccines but was mostly abandoned by 2001 as a precautionary measure. RotaTeq doesn’t and never did contain thimerosal—because it’s a live, attenuated viral vaccine, it doesn’t contain any preservatives. But Kennedy has frequently used thimerosal as a vaccine bogeyman over the years, claiming it causes harms (there is no evidence for this).

After their conversation, Kennedy published a story in Rolling Stone and Salon.com titled “Deadly Immunity,” which erroneously argued that thimerosal-containing vaccines cause autism. The article was riddled with falsehoods and misleading statements. It described Offit as “in the pocket” of the pharmaceutical industry and claimed RotaTeq was “laced” with thimerosal. Rolling Stone and Salon amended some of the article’s problems, but eventually Salon retracted it and Rolling Stone deleted it.

Looking back, Offit said he was sandbagged. “He’s a liar. He lied about who he was; he lied about what he was doing. He was just wanting to set me up,” Offit said.

Although that was the only time they had ever spoken, Kennedy has continued to disparage and malign Offit over the years. In his book dedicated to denigrating Fauci, Kennedy spends plenty of time spitting insults at Offit, calling him a “font of wild industry ballyhoo, prevarication, and outright fraud.” He also makes the wildly false claim that RotaTeq “almost certainly kills and injures more children in the United States than the rotavirus disease.”

Inconvincible

Understanding that Kennedy is a germ theory denialist and terrain theory embracer makes these attacks easier to understand—though no less abhorrent or dangerous.

“He holds these beliefs like a religious conviction,” Offit said. “There is no shaking him from that,” regardless of how much evidence there is to prove him wrong. “If you’re trying to talk him out of something that he holds with a religious conviction—that’s never going to happen. And so any time anybody disagrees with him, he goes, ‘Well, of course, they’re just in the pocket of industry; that’s why they say that.'”

There are some aspects of terrain theory that do have a basis in reality. Certainly, underlying medical conditions—which could be considered a disturbed bodily “terrain”—can make people more vulnerable to disease. And, with recent advances in understanding the microbiome, it has become clear that imbalances in the microbial communities in our gastrointestinal tracts can also predispose people to infections.

But, on the whole, the evidence against terrain theory is obvious and all around us. Terrain theorists consider disease a symptom of an unhealthy internal state, suggesting that anyone who gets sick is unhealthy and that all disease-causing germs are purely opportunistic. This is nonsense: Plenty of people fall ill while being otherwise healthy. And many germs are dedicated pathogens, with evolved, specialized virulence strategies such as toxins, and advanced defense mechanisms such as antibacterial resistance. They are not opportunists.

(There are some terrain theory devotees who do not believe in the existence of microbes at all—but Kennedy seems to accept that bacteria and viruses are real.)

Terrain theory applied

Terrain theory’s clash with reality has become painfully apparent amid Kennedy’s handling—or more accurately, mishandling—of the current measles situation in the US.

Most health experts would consider the current measles situation in the US akin to a five-alarm fire. An outbreak that began at the end of January in West Texas is now the largest and deadliest the country has seen in a quarter-century. Three people have died, including two unvaccinated young children who were otherwise healthy. The outbreak has spread to at least three other states, which also have undervaccinated communities where the virus can thrive. There’s no sign of the outbreak slowing, and the nation’s overall case count is on track to be the highest since the mid-1990s, before measles was declared eliminated in 2000. Modeling indicates the country will lose its elimination status and that measles will once again become endemic in the US.

Given the situation, one might expect a vigorous federal response—one dominated by strong and clear promotion of the highly effective, safe measles vaccine. But of course, that’s not the case.

“When those first two little girls died of measles in West Texas, he said immediately—RFK Jr.—that they were malnourished. It was the doctors that stood up and said ‘No, they had no risk factors. They were perfectly well-nourished,'” Offit points out.

Kennedy has also heavily pushed the use of vitamin A, a fat-soluble vitamin that accumulates in the body and can become toxic with large or prolonged doses. It does not prevent measles and is mainly used as supportive care for measles in low-income countries where vitamin A deficiency is common. Nevertheless, vaccine-hesitant communities in Texas have embraced it, leading to reports from doctors that they have had to treat children for vitamin A toxicity.

Poisons

Despite the raging outbreak, Kennedy spent part of last week drumming up fanfare for a rickety plan to rid American foods of artificial food dyes, which are accused of making sugary processed foods more appealing to kids, in addition to posing their own health risks. It’s part of his larger effort to improve Americans’ nutrition, a tenet of terrain theory. Though Kennedy has organized zero news briefings on the measles outbreak, he appeared at a jubilant press conference on removing the dyes.

The conference was complete with remarks from people who seem to share similar beliefs as Kennedy, including famed pseudoscience-peddler Vani Hari, aka “Food Babe,” and alternative-medicine guru and fad diet promoter Mark Hyman. Wellness mogul and special government employee Cally Meads also took to the podium to give a fury-filled speech in which he claimed that 90 percent of FDA’s spending is because we are “poisoning our children,” echoing a claim Kennedy has also made.

Kennedy, for his part, declared that “sugar is poison,” though he acknowledged that the FDA can’t ban it. While the conference was intended to celebrate the removal of artificial food dyes, he also acknowledged that there is no ban, nor forthcoming regulations, or even an agreement with food companies to remove the dyes. Kennedy instead said he simply had “an understanding” with food companies. FDA Commissioner Marty Makary explained the plan by saying: “I believe in love, and let’s start in a friendly way and see if we can do this without any statutory or regulatory changes.” Bloomberg reported the next day that food industry lobbyists said there is no agreement to remove the dyes.

However feeble the move, a focus on banning colorful cereal during a grave infectious disease outbreak makes a lot of sense if you know that Kennedy is a germ theory denialist.

But then again, there’s also the brain worm.

Photo of Beth Mole

Beth is Ars Technica’s Senior Health Reporter. Beth has a Ph.D. in microbiology from the University of North Carolina at Chapel Hill and attended the Science Communication program at the University of California, Santa Cruz. She specializes in covering infectious diseases, public health, and microbes.

RFK Jr. rejects cornerstone of health science: Germ theory Read More »

the-end-of-an-ai-that-shocked-the-world:-openai-retires-gpt-4

The end of an AI that shocked the world: OpenAI retires GPT-4

One of the most influential—and by some counts, notorious—AI models yet released will soon fade into history. OpenAI announced on April 10 that GPT-4 will be “fully replaced” by GPT-4o in ChatGPT at the end of April, bringing a public-facing end to the model that accelerated a global AI race when it launched in March 2023.

“Effective April 30, 2025, GPT-4 will be retired from ChatGPT and fully replaced by GPT-4o,” OpenAI wrote in its April 10 changelog for ChatGPT. While ChatGPT users will no longer be able to chat with the older AI model, the company added that “GPT-4 will still be available in the API,” providing some reassurance to developers who might still be using the older model for various tasks.

The retirement marks the end of an era that began on March 14, 2023, when GPT-4 demonstrated capabilities that shocked some observers: reportedly scoring at the 90th percentile on the Uniform Bar Exam, acing AP tests, and solving complex reasoning problems that stumped previous models. Its release created a wave of immense hype—and existential panic—about AI’s ability to imitate human communication and composition.

A screenshot of GPT-4's introduction to ChatGPT Plus customers from March 14, 2023.

A screenshot of GPT-4’s introduction to ChatGPT Plus customers from March 14, 2023. Credit: Benj Edwards / Ars Technica

While ChatGPT launched in November 2022 with GPT-3.5 under the hood, GPT-4 took AI language models to a new level of sophistication, and it was a massive undertaking to create. It combined data scraped from the vast corpus of human knowledge into a set of neural networks rumored to weigh in at a combined total of 1.76 trillion parameters, which are the numerical values that hold the data within the model.

Along the way, the model reportedly cost more than $100 million to train, according to comments by OpenAI CEO Sam Altman, and required vast computational resources to develop. Training the model may have involved over 20,000 high-end GPUs working in concert—an expense few organizations besides OpenAI and its primary backer, Microsoft, could afford.

Industry reactions, safety concerns, and regulatory responses

Curiously, GPT-4’s impact began before OpenAI’s official announcement. In February 2023, Microsoft integrated its own early version of the GPT-4 model into its Bing search engine, creating a chatbot that sparked controversy when it tried to convince Kevin Roose of The New York Times to leave his wife and when it “lost its mind” in response to an Ars Technica article.

The end of an AI that shocked the world: OpenAI retires GPT-4 Read More »