Memory maker CXMT has waded into the Apple memory sourcing debate, suing the Pentagon for designating it as a risky “Chinese military company.”

In June, Apple petitioned the Trump administration to allow it to buy memory chips from a blacklisted Chinese supplier. Now, that same supplier is bringing the fight to the U.S. government.

Memory chip producer CXMT sued the Pentagon on Friday, reports Reuters, for including it on the Chinese Military Company Blacklist, or 1260H list.

It’s a list of companies that the Pentagon believes have links to China’s People’s Liberation Army. The companies on it are deemed to potentially be a risk to U.S. national security.

CXMT insists that it is not a military company with no affiliation with the country’s army. It says it designs chips for civilian and commercial use.

The Pentagon, meanwhile, made its decision arbitrarily without any real supporting evidence. As a result, CXMT says says that it has damaged its business and reputation.

Getting off the list

CXMT has been classified as a risk by the Pentagon since January 2025, with the designation maintained in June 2026 under the Trump administration.

However, CXMT has spent over a year providing information to the Defense Department to try and convince it the classification is unjustified.

In February 2026, there seemed to be progress, as the Pentagon briefly said CXMT would be taken off the list, the company claims. However, later that day, the notice was pulled with no explanation.

In June, the Pentagon called CXMT “directly affiliated with MIIT and indirectly affiliated with SASAC,” referring to the Ministry of Industry and Information Technology as well as the China-controlled Assets Supervision and Administration Commission.

A bad list

The function of the list is to dissuade U.S. companies from buying goods from listed firms. The list doesn’t specifically forbid purchases from CXMT, but there are consequences.

The main repercussion is that the Department of Defense cannot make agreements with companies that exist on the list. However, it also won’t use products or services from third parties that use those components.

If Apple sells goods using the memory, it would endanger other sales to the U.S. government.

It’s also not the worst list, as there’s also the Entity List. Companies on the Entity List are blocked from trade with all companies in the U.S.

For Apple, if it proceeded to use CXMT, there would be a risk of the memory supplier being included on the Entity List, scuppering Apple’s new supply source.

Apple’s request to use CXMT as a supplier has so far been a political issue. It’s had lot of pushback on the topic from Washington as well as rival memory supplier Micron.

CXMT’s legal challenge forces the issue, and could make the whole debate conclude on a much faster schedule.

What just happened? A federal appeals court ruled on Friday that Nevada can enforce its gambling laws against Kalshi’s sports prediction markets, dealing the company a major setback in its fight to keep state regulators out of the business. The decision could give regulators across the West more leverage to challenge sports contracts offered on federally regulated prediction exchanges, while moving the industry closer to a Supreme Court fight over whether those products are financial derivatives or gambling.

The Ninth Circuit Court of Appeals found that Kalshi’s sports contracts amount to sports betting, even though the company offers them through a federally regulated exchange. The decision could have implications well beyond Nevada, particularly in states that have been trying to limit or block sports-related prediction markets.

Kalshi lets users trade contracts based on event outcomes. The platform lists markets tied to sports, elections, economic data, weather, entertainment, and other subjects. Its sports contracts have become a major part of its activity as prediction markets have grown quickly this year.

The company has argued that its contracts are federally regulated financial products, not gambling. Kalshi is regulated by the Commodity Futures Trading Commission and has said that federal oversight should prevent states from applying their own gaming laws to its operations.

The Ninth Circuit rejected that argument.

“The substance of the sports event contracts offered on Kalshi’s (exchange) is sports gambling, regardless of whether Kalshi calls them swaps,” the three-judge panel wrote.

The court also said Kalshi had not shown a meaningful difference between its sports markets and traditional sportsbook wagers. “Kalshi’s attempts to distinguish its sports event contracts from sportsbooks betting are unpersuasive,” the ruling said.

The panel took particular issue with Kalshi’s position that its products were not sports betting. The judges said it was “disingenuous” for the company to make that argument after using sports-betting language in earlier marketing materials.

The case began after Nevada regulators moved against Kalshi and similar prediction-market offerings from Crypto.com and Robinhood. Nevada has long treated sports wagering as a tightly regulated activity, and state officials argued that the platforms should be subject to the same rules and taxes as other betting operators.

Gov. Joe Lombardo said the ruling would help “safeguard the integrity of our gaming industry.” Mike Dreitzer, chair of the Nevada Gaming Control Board, said it “completely vindicates what we have been saying all along.”

Kalshi said it intends to challenge the ruling. Spokeswoman Dani Lever said the company believes CFTC regulations do not bar sports contracts and noted that the agency is considering further clarification of its rules.

“Despite the Ninth Circuit’s opinion, we still believe the CFTC regulations as written do not prohibit sports contracts, and in any event, the CFTC is working to clarify those regulations,” Lever said in a statement. “We will be seeking further review.”

The decision conflicts with a ruling earlier this year from the Third Circuit Court of Appeals. In that case, the court blocked New Jersey from enforcing its gaming laws against Kalshi and other prediction-market companies.

That split raises the odds that the Supreme Court will eventually decide whether federally regulated prediction exchanges can offer sports contracts without state gaming licenses.

“It’s the first ruling against Kalshi at the appellate level, and the opinion seemed to be pretty brutal for the company,” Dustin Gouker, an independent journalist who covers the prediction industry, told CNN. “This gets us one step closer to an almost inevitable Supreme Court case on the legality of sports event contracts.”

The Ninth Circuit’s ruling now applies to federal courts in California, Arizona and six other states in the circuit. It could give regulators in those states more leverage as they challenge prediction-market companies.

The ruling comes as the industry is gaining users and trades. Kalshi has reported billions of dollars in weekly trading volume, with sports markets driving much of that activity. Polymarket, the second-largest prediction platform in the United States, was not part of the Nevada case, but the legal issues raised by the ruling could affect its business as well.

ai and ml

Company’s three year pursuit of debt from non-existent address ended by Oxford judge

British energy company SSE Energy Supply refused to believe that there was no unit 8b at the property of Lyle Hopkins, a doctoral student at the University of Oxford.

For more than 20 months, the company billed Hopkins at business rates, a total of £1,091.01, for an unused electricity meter on the property that had been associated with a previous owner’s unsuccessful effort to convert the property into flats.

A company representative even sent an email on June 15, 2024, stating that Hopkins was not liable. Nonetheless, SSE then sent in debt collectors to recover debt attributed to a disconnected meter for a non-existent address.

He never paid the bill. But he wanted the harassment to end.

So Hopkins, a software engineer working on his PhD at Oxford Interdisciplinary Bioscience DTP, turned to GPT-5.5 and Claude Fable for legal guidance to challenge the unwarranted and annoying collection effort. Representing himself in court, he prevailed on July 17, when a judge at the County Court at Oxford (St Aldate) awarded him £1,087.88, including expenses and interest.

Citing The Register’s coverage of how courts have required lawyers to declare and verify AI-generated documents, Hopkins said his experience has been that it’s the verification that matters.

“I used AI to test an energy company’s court claims against its own records, then ran the hearing myself and won,” he said.

Hopkins said he wouldn’t have been able to quote the case law without the help of AI.

“I had to review everything and give it steering, and make sure the case law actually existed,” he said. “I mostly just gave it the facts and it came up with the arguments, checked the laws, court rules, case law, and regulations. Really I just knew what they had done couldn’t be right, and relied on the AI to find the legal justifications as to why.”

Hopkins said that when he first submitted his claim back on October 7, 2025, GPT models hallucinated more and weren’t as good at checking case law.

“If Fable had been around at the time, I probably would have claimed for a lot more, probably around £5,000 and included harassment, which would have been more in line with the case law it found when we were preparing the court documents about three weeks before the hearing.”

Hopkins didn’t consider a solicitor because the cost would have been prohibitive. His spending on AI was more modest but not trivial – three active AI subscriptions and extra API fees.

“I burned a lot of tokens on this, on my Github Copilot subscription, OpenAI subscription, and Anthropic subscription,” he said. “Due to the deadlines I ended up paying API rates for some of it, which was painful.” (~£175).

His award could have been higher. “Before the hearing, SSE offered me more than the court ultimately awarded, conditional on confidentiality and non-disparagement clauses,” Hopkins said. “I refused because it was more important to be able to tell people what they had done.”

A rollercoaster ride

The judge hearing the case – shared with The Register – said Hopkins, who sought the intercession of an Energy Ombudsman after repeated communication with SSE, could not have done more to make clear that he was not liable for claimed debt.

“The defendant has subjected Mr Hopkins to a rollercoaster ride and not a good one,” the judge said, adding, “pursuing Mr Hopkins for that liability, if continued, would amount to harassment.”

Yet even after Hopkins won, the billing contniued. Following the judgement on July 17, another bill  datedJuly 23 went out.

“Since the judgment…in direct contradiction to the ruling which stated any further attempts to claim the invalid debt from me would constitute harassment, they have sent me another bill with payment demand,” Hopkins told The Register.

He’s unsure whether he will pursue a harassment claim.

“This dragged on for nearly three years and cost hundreds of hours, lost sleep, and time that should have gone into my medical-research doctorate at Oxford,” he said. “My exhibits alone were over 100 pages. Given how difficult this was, I struggle to see how people less academic than me wouldn’t end up with CCJs and ruined credit ratings over debts they never owed.”

Hopkin’s doctoral funding ended before he could finish his degree, a delay he attributes to time spent on this case. So he has launched a funding campaign to support the completion of his doctorate. 

The Register emailed SSE seeking comment and received an autoreply indicating that that SSE media team was unavailable outside of business hours in the UK. The urgent inquiry number yielded a representative who indicated that he only handled questions about outages. If company representatives reply after returning to the office, we’ll update this story. ®

• Microsoft has made a welcome change to OneDrive in Windows 11
• Prompts to use the backup service can now be permanently dismissed
• The change has been positively received by most users online

There have been complaints about this going back years, but Microsoft has finally caved in and decided to stop nagging users to back up their files to OneDrive: these alerts can now be permanently dismissed with an Opt out of backup button.

As spotted by Windows Latest, the move is going to please users who have no intention of using OneDrive to store their files, and would rather not have occasional reminders from their operating system that this is an option.

Microsoft has spent the past few months testing a much cleaner version of Windows Search, and it is finally starting to reach regular Windows 11 PCs.

According to Windows Latest, the redesigned Search experience is included in the August 2026 optional update, KB5120998, for Windows 11 24H2 and 25H2. Microsoft confirms that the update moves PCs to builds 26100.9278 and 26200.9278, respectively, although some of its new features are being enabled gradually.

We first covered these Search changes in July, when they were limited to Windows Insiders in the Experimental channel.

You can finally keep Bing out of Windows Search

The biggest improvement is a pair of controls under Settings > Privacy & security > Search. One toggle controls web results, while another handles Microsoft Store suggestions.

Turning off web results removes Bing suggestions from Search, while disabling both leaves Windows focused almost entirely on results from the PC. Windows Latest also found that Search feels faster this way because it no longer waits for an online Bing request before returning results.

The Search home screen is cleaner as well. Microsoft has removed much of the old Bing-powered clutter, including trending searches and promotional content, while results now clearly identify whether they come from an app, file, setting, the web, or Microsoft Store. Typo handling and local result prioritization have also been improved.

The rollout is still gradual

Installing KB5120998 does not necessarily mean the new Search will appear immediately. Microsoft is using its Controlled Feature Rollout system, meaning the necessary components can already be installed while the feature itself remains disabled on some PCs.

Windows Latest says users who do not want to wait can force-enable it through ViVeTool, although everyone else can simply wait for Microsoft to activate the feature automatically.

Windows Search should never have needed a web request to help someone find Calculator, Settings, or a file sitting on their own PC. Microsoft has taken its time getting here, but the option to remove Bing finally makes Search behave much closer to what a desktop search tool should be.

In an age when every click counts and digital footprints are incessantly monitored, maintaining true privacy online feels increasingly like an elusive luxury. Most people who browse the internet are, all the time, being identified by one digital address: their IP address. At its simplest level, this address communicates to the world (or, at least, what websites you visit) exactly where you are and who you could be.

It’s a little more than the digital equivalent of a name tag that broadcasts your home address, too. What now, though, is a smart and rapidly emerging answer that empowers users to regain anonymity and look assuredly: residential proxies.

In short, a residential proxy is a middleman connecting homeowners via an actual, physical IP address allocated by an Internet Service Provider (ISP) to an actual homeowner. Your connection looks to be one that comes from an average house, not a corporate data centre, when you use one. That distinction is also important for privacy and access.

Real IP Addresses and the Advantages of Them

Privacy online is a lot more than hiding your location. That means regulating the information companies capture about your behaviours. Every search query, every product you see and every news article you read under that single, recognisable marker gets logged when you open your native IP. This data is then aggregated to make a full profile of you that is used with targeted advertising, price discrimination and content tailoring.

A residential proxy serves as an advanced digital disguise. By cycling over and over again, through a range of thousands or even millions of valid, unique residential IPs, you are essentially splintering your digital identity. In the opinion of an outsider, your action almost seems to have been performed by thousands or millions of different people over a period of time. This rotation means no one thing can really create an all-around accurate overview of your browsing activity. This rotation is what renders them a simple and efficient form of remaining truly private.

Upholding the Consumer Advantage: Unbiased Access

This newfound privacy is vital for numerous areas and functions, the most obvious of which is market research and price comparison. Let’s say you want to see the price of a flight or a product. Companies commonly use dynamic pricing where they reflect higher prices to those they think are wealthier, more frequent or visiting from a particular place. If you visit the website using your home IP, they already know your history.

If you visit it by plugging into a residential proxy from another city or region, the site is fooled into believing you are an entirely new, unique visitor, so you can see unadulterated, normal pricing. In doing so, it transforms the technology from an industrial tool of surveillance to a tool of customer empowerment and promotes a more equitable online marketplace.

Stability Means Simplicity

Further, the convenience or convenience of residential proxies is dependable and easy to integrate. The most serious problem faced by online work in recent years has been constant safety warnings, CAPTCHA issues or soft-bans. For low-quality and easily visible IPs, your work grinds to a halt.

Residential IPs face fewer such challenges because they are genuine. The connection is seen as a normal and common request, which helps for a smooth, consistent experience, and less downtime and red tape. And that stability results directly in more efficiency and peace of mind — for everyone dependent on the internet to accomplish complex tasks, from handling multiple social media accounts to gathering information on the web.

In addition, setting up and using these proxies is also not as problematic as we think. Modern proxy services present the technology to all, not just technologists. There’s a simple interface that you’ll be able to configure and work out pretty easily on your own — you pay for a service, install a small application or add a browser function, and it starts instantly routing traffic through a bunch of residential IP networks.

The proxy manager manages the hard work of changing addresses, location requests and speed of connections. At the end of the day, the end users are free to browse the internet the way they wish when it suits them — privately and safely.

Finally, no privacy was built into the internet. Nowadays, people need to actively protect the people who take care of their information and identity. Residential proxies are a robust but uncomplicated solution.

These proxies hide your real and personal locations behind a real IP address of a real home user, providing your anonymity, keeping you hidden from aggressive tracking and surveillance, providing objective information that’s useful and complex, and performing the tasks of using the internet without ever having to be disturbed. It’s not just an expert tool, but the key to making the most of the true online world of the new normal. 

Eighteen months after their merger fell apart over who would control whom, Honda and Nissan have agreed to share the software inside their cars.

The two signed a joint development agreement on Monday covering electronic control units, an in-vehicle operating system, middleware and vehicle control software, with the first cars arriving from fiscal 2029.

The agreement is narrower than a merger and, in engineering terms, more intimate. Standardising core ECUs means agreeing on the computers that run the brakes, the steering, the battery and the cabin, and on the interfaces between them.

Both companies framed the decision as a matter of pace and cost. Faster development cycles and more efficient investment were the stated reasons, after what they described as extensive studies of where collaboration in the software domain might work.

Neither disclosed what it will cost or save. No supplier was named, no executive was quoted by name, and the specific ECUs to be standardised were not identified.

Fiscal 2029 is the part worth sitting with. That is roughly three years from now for architecture work that has to be locked long before the first vehicle reaches a showroom, which tells you how far ahead this kind of decision is made and how little room there is to change it afterwards.

The pieces being standardised are the ones customers never see and never forgive. An operating system that stutters, an over-the-air update that fails, a control unit that disagrees with another: these are the faults that define whether a car feels modern.

The problem they are addressing is real and not specific to Japan. A modern car runs on tens of millions of lines of code across dozens of controllers, and every manufacturer is discovering that writing all of it alone is expensive in a way that engines never were.

Scale is the only available answer. Arm has been arguing that the industry needs common foundations rather than dozens of bespoke stacks, which is precisely what two carmakers sharing an operating system amounts to.

The competitive pressure is coming from a specific direction. BYD has built its own 4nm driving chip and is fitting LiDAR to a $10,000 car, which is a level of vertical integration neither Japanese company can match alone.

Chinese brands have been taking share in Europe at a record rate, and the software experience inside the vehicle is a large part of why. Buyers comparing a Chinese EV with a Japanese one in 2029 will be comparing operating systems as much as drivetrains.

What makes this agreement notable is the history behind it. Merger talks collapsed in February 2025 after Honda proposed making Nissan a wholly owned subsidiary, an arrangement Nissan would not accept, and the two walked away.

Sharing an operating system without sharing ownership is a considerably harder thing to sustain. Every future disagreement about a feature, a release schedule or a security patch now has to be resolved between two independent boards with different product cycles and different problems.

There is a precedent both will know well. Shared platforms have worked in this industry for decades, but they involved metal and tooling with fixed specifications, whereas software has to keep being decided long after the vehicle ships.

Nissan has the more urgent problems of the two, having spent the period since the merger collapse restructuring. Honda is the larger and steadier partner, which is the imbalance that sank the merger and has not gone anywhere.

The pair have said they will keep exploring collaboration elsewhere, naming carbon neutrality and the elimination of traffic fatalities as areas of shared interest. Those are aspirations rather than agreements, and the distinction is doing work in the announcement.

What exists today is a signed document, a defined technical scope, and a date in 2029. Whether it holds together until then is a question about two organisations rather than about software.

You’ll need your old passcode or Apple Account password.

How to fix an iPhone Unavailable message

Tatiana Diuvbanova/Shutterstock

Nearly every company that’s trying to deploy artificial intelligence runs into the same problem: it’s hard to integrate the tech into everyday operations. Industrial heavyweight Caterpillar has spent decades dealing with a version of that problem in the physical world, and now it’s using its experience to deploy AI.

Caterpillar’s push into the autonomous space started with mining, where labor shortages and hazardous conditions can make automation particularly useful. Today, it sells automated haul trucks, drilling, underground loaders, dozers, remote-controlled construction equipment, and more. It also offers a software command center, fleet management, and even remote terrain intelligence as part of its autonomous toolkit.

“Now we’re in this super exciting time where we can take all of that learning from mining and bring it into much more dynamic environments, jobsites, quarries, and construction sites,” the company’s CTO, Jaime Mineart, told TechCrunch on during a fireside chat at the Ai4 conference in Las Vegas earlier this month.

The industrial giant is now applying AI more broadly, including in tools used by technicians and its own employees. One example is the Cat AI Assistant, which lets field technicians standing next to a machine use voice commands to pull up repair procedures, troubleshoot potential problems, and identify parts that may be needed before beginning a repair. Mineart said the tool is now being used by customers, operators and technicians.

The assistant draws on Caterpillar’s proprietary data, which spans information generated by its connected machines. Mineart said Caterpillar has about 1.6 million connected assets globally and more than 16 petabytes of structured data.

The company is also using AI to power software for scanning sites and generating digital twins in manufacturing to analyze operations, she said. And like nearly every other company, Caterpillar is using AI across its enterprise operations, as well as for software development. “We use AI agents to modernize legacy code, generate and test new software, and identify defects earlier,” Mineart said.

But Mineart is quick to point out that building the technology is only part of the challenge, as deploying an autonomous machine is not the same as transforming a site to use AI. Companies also have to rethink how people work alongside the technology and how existing processes need to change.

“The hard part about autonomy and about physical AI is incorporating that technology into the customer jobsite and into the workflows,” she said.

Mineart said the company leans on experienced operators to help train AI systems, leveraging institutional knowledge built over decades. And as machines become more autonomous, some operators may shift from controlling a single machine to overseeing multiple machines from a remote command center.

That transition, however, is creating a new challenge for Caterpillar: training its 118,000 employees. Mineart said the company plans to spend $100 million over the next five years to train its workforce in AI, autonomy and robotics.

That investment is likely being put towards helping the company make the most of the broader boom in AI infrastructure, which is already helping its top-line. Caterpillar’s quarterly revenue reached an all-time high of $20.5 billion in the second quarter, helped by strong demand for power-generation equipment used in data centers. Its power-generation division saw sales spike 72% to $3.10 billion, and CEO Joe Creed said that “no one is slowing down” when it comes to demand for cloud computing and generative AI infrastructure.