A bipartisan group of UK politicians is sounding the alarm over the country’s partnership with the data analytics company Palantir.

In a report published Tuesday, the 11 members of Parliament’s Science, Innovation, and Technology Committee warned that the country’s ballooning reliance on Palantir’s technology “represents an unacceptable point of weakness” that could hand the company overwhelming bargaining power in future negotiations.

“We know that with vendor lock-in, over time, we’ll get more expensive and worse services,” Dame Chi Onwurah, chair of the committee and member of Parliament, tells WIRED. “It’s a trap that has to be avoided.”

In a worst-case scenario, a deeply entrenched supplier could threaten to withhold service as a way of imposing its will, Onwurah believes. “That could bring public services and our economy to a halt,” she says. “That’s a huge risk.”

Though the committee says that its objections to Palantir are not ideologically motivated, the report also describes a “clear mismatch with UK values.” It points to politically charged comments by Palantir cofounder Peter Thiel—who in 2023 described the British public’s affection for the NHS as “Stockholm syndrome”—and a 22-point manifesto based on a recent book by CEO Alex Karp, which advocates for an overriding fealty to the US and its interests.

“We have a key vendor saying they will exercise technology in accordance with their political mission,” Onwurah says. “If what the UK is trying to do in our NHS or our defense does not align with Palantir’s political objectives, we clearly can’t depend upon them as a supplier.”

To minimize the risks, the committee recommended that the National Health Service, one of Palantir’s primary partners in the UK, activates a clause in its contract next February that would terminate the relationship early.

The UK government began to use Palantir’s technology in 2020 as it scrambled to map the spread of the Covid-19 virus and route medical equipment across the country. Since then, Palantir and its partners have won contracts worth a combined $750 million with the NHS and the Ministry of Defense, among others. The company has touted its ability to enable “innovation and fast-paced problem solving” in the UK public sector.

The report outlines similar dependencies on US-based cloud providers Microsoft and Amazon Web Services, and Fujitsu, the Japanese company at the center of the Post Office Horizon scandal. But “Palantir concerns us most,” the committee wrote.

Palantir did not immediately respond to a request for comment.

The relationship has attracted increased scrutiny of late over the company’s work with US Immigration and Customs Enforcement (ICE), as well as the US and Israeli militaries. The manifesto based on Karp’s book further inflamed concerns about the company’s politics.

“They’re not a company that should be anywhere near British public services,” says Donald Campbell, director of advocacy at Foxglove, a nonprofit that has previously campaigned for the NHS to back out of its contract with Palantir. “Do you want to be giving a company of this kind—with these openly expressed opinions and ideologies—a central role in the UK state that it may get harder and harder to remove them from?”

Appearing before the committee in July last year, Louis Mosley, who heads up Palantir’s European business, distanced the company from Thiel’s comments about the NHS. Palantir’s objective is to “support democratically elected governments in delivering the mandate that they have been elected to deliver,” he said. “We represent a diversity of political views and do not take political positions as a company.”

With little fanfare, US President Donald Trump may have signed one of the most important executive orders in his second term at the White House. With the “Promoting Advanced Artificial Intelligence Innovation and Security” EO, the US government is finally putting its finger on the scale of AI development, more or less demanding that AI companies provide it with access to their Frontier Models for a period of 30 days before their public release.

Since no national US regulations currently exist for AI and much of the oversight is being left to a hodgepodge of mostly in-process state-level regulation, this is the first whiff of broad-based control.

I’ve been reviewing robot vacuum cleaners since Ecovacs Robotics brought its first Deebot to Australia about a decade ago. So I like to think that I’ve learnt a few things about what to look for when buying one — knowledge that I’ve used to curate the best robot vacuums in Australia right now.

The manufacturers don’t make it easy, do they? They make every spec sheet sound glowing, but you get what you pay for, especially when it comes to robot vacuums. It’s understandable that you’d want an affordable model, and there are plenty that make sacrifices to keep costs down — you just need to be able to understand what those sacrifices mean for your personal use case.

In the same vein, you don’t have to spend thousands on a robovac, although you will get the top-of-the-line models at premium prices that offer plenty of smart features. Some of those, like agentic voice assistants and Matter connectivity, aren’t really necessary, but you might want pet-specific features that cheaper models won’t have. But not all expensive models are reliable cleaners.

So, what’s a good robot vacuum to buy then? I’ve taken a look at the dozens of robovac discounts on Amazon‘s Mid-Year Sale and picked four at different price points even though they all do pretty much the same thing — vacuum, mop, and clean themselves. Without a discount, they’re all on the expensive side, but the offers make each of them a good-value proposition for different reasons.

Follow my live coverage of Amazon’s Mid-Year Sale for more vacuum cleaner discounts.

My top robovac deals

What to look for in a robot vacuum

Looking for the most recent regular Connections answers? Click here for today’s Connections hints, as well as our daily answers and hints for The New York Times Mini Crossword, Wordle and Strands puzzles.

I have a complaint about the yellow group in today’s Connections: Sports Edition puzzle: How can a word be a synonym for itself? That will make sense once you play today’s game. If you’re struggling with the puzzle but still want to solve it, read on for hints and the answers.

Connections: Sports Edition is published by The Athletic, the subscription-based sports journalism site owned by The Times. It doesn’t appear in the NYT Games app, but it does in The Athletic’s own app. Or you can play it for free online.

Read more: NYT Connections: Sports Edition Puzzle Comes Out of Beta

Hints for today’s Connections: Sports Edition groups

Here are four hints for the groupings in today’s Connections: Sports Edition puzzle, ranked from the easiest yellow group to the tough (and sometimes bizarre) purple group.

Yellow group hint: Helps to be tall.

Green group hint: They’re used to winning.

Blue group hint: Hoops players.

Purple group hint: Clues relating to one player.

Answers for today’s Connections: Sports Edition groups

Yellow group: Dunking synonyms.

Green group: Teams with 5+ NBA titles.

Blue group: Nicknames of players in the NBA finals.

Purple group: Associated with Jalen Brunson.

Read more: Wordle Cheat Sheet: Here Are the Most Popular Letters Used in English Words

What are today’s Connections: Sports Edition answers?

The yellow words in today’s Connections

The theme is dunking synonyms. The four answers are dunk, jam, slam and stuff.

The green words in today’s Connections

The theme is teams with 5+ NBA titles. The four answers are Bulls, Celtics, Lakers and Spurs.

The blue words in today’s Connections

The theme is nicknames of players in the NBA finals. The four answers are Deuce, Kat, Swipa and Wemby.

The purple words in today’s Connections

The theme is associated with Jalen Brunson. The four answers are 11, ECF MVP, Knicks and Villanova.

When hackers break into your computer, one of the first things they do is delete or lock your files and vanish. By the time you realize something is wrong, the damage is often permanent. A researcher at Florida International University has found a way to change that, and the solution is built right into your storage drive.

Understanding where your deleted files actually go

When you delete a file, it does not disappear immediately. It sits in a kind of digital purgatory, existing in fragments on your drive before being permanently wiped to free up space.

If a hacker encrypted or erased your data, you could theoretically reach back in and pull those files out before they vanish for good. The problem is that modern SSDs, which power most laptops and computers today, manage this space carelessly.

When the drive needs space, it clears deleted data based purely on efficiency, with no awareness of how recently files were removed. Files deleted during a ransomware attack could get wiped first, while old junk files from weeks ago survive.

How your SSD can be turned into a cybersecurity tool

The system works by sequencing deleted data by age, so the oldest deleted files go first, and the most recently deleted files stay protected for as long as possible. It also extends the window for recovering deleted data to up to 126 days, improving data protection by at least 60% with minimal impact on drive speed.

Since the storage drive operates independently from your operating system, it can keep protecting your data even after hackers have taken full control of your software. The research is now in active discussions with industry partners about bringing the technology to market.

If this research has you thinking about what your storage drive is really up to, you might also want to read about how websites can spy on your browsing habits through your hard drive.

cyber-crime

You don’t infect anyone in Russia or other CIS countries

Even ransomware cartels make mistakes, and in this case, it was a biggie that could have landed the responsible crim in a Russian gulag: accidentally infecting a company located in a Commonwealth of Independent States country.

In what threat-hunter Dominic Alvieri deemed the ransom “dumbass of the day,” Nova, the affiliate program for ransomware crew RAlord, on Tuesday issued an apology to Eriell Group, a major oilfield services company with headquarters in Uzbekistan and a corporate office in Moscow.

Apparently, Eriell contacted Nova and notified the ransomware operators about an affiliate’s mess-up.

The affiliate has since been banned from the criminal operation, we’re told. In addition to issuing a “formal apology,” the ransomware gang promised to assist Eriell with the recovery process “free of charge.” The malware slingers claimed they didn’t encrypt any files, and pledged not to leak any of the stolen data.

“Apparently, the first rule of ransomware club, you don’t attack organizations in the Commonwealth of Independent States (CIS), is still very much in effect in 2026,” Recorded Future threat intelligence analyst Allan Liska told The Register.

While cybercrime is technically illegal in Russia and other CIS countries, their governments often provide safe harbor for extortionists and other financially motivated crims – especially if they also happen to work day jobs as state-sponsored hackers – and local police look the other way unless the gangs infect any in-country organizations.

Some crews, like the DragonForce cartel, VanHelsing ransomware-as-a-service group, and notorious LockBit operators, expressly prohibit their gang members and affiliates from hitting Russian and other CIS targets.

We’re guessing that the Nova affiliate will be high up on all of these gangs’ do-not-hire lists for quite a while.

Still, they aren’t the first cybercriminal, Russian-speaking or otherwise, to make seriously dumb mistakes.

Earlier this year, notorious data-leak-and-extortion crew Scattered Lapsus$ Hunters claimed they had gained “full access” to Resecurity’s systems and stolen “everything.” Resecurity later offered its “congratulations” to the cybercrime crew, which had fallen into the threat intel team’s honeypot – resulting in a subpoena being issued for one of the data thieves. 

Pro-Russian hacktivist crew CyberVolk got sloppy when they debuted a ransomware service late last year. They hardcoded the master keys – this same key encrypted all files on a victim’s system – into the executable files, thus allowing victims to recover encrypted data without paying any extortion fees.

While that mess-up worked in the victim orgs’ favor, another coding error committed by Sicarii malware developers makes it nearly impossible for companies to recover their files: the Sicarii encryptor generates a new cryptographic key pair during every execution – but then discards the private key, meaning there’s no recoverable master key.

Similarly, a programming mistake in Nitrogen ransomware prevents the gang’s decryptor from recovering victims’ files, again making paying up futile.

Trellix VP of threat intel strategy John Fokker recently told us that he got so sick of seeing the security industry “glorifying threat actors,” that he and his team decided to troll the baddies, and started publishing the Dark Web Roast.

“These are just individuals, they just use computers, and they just want to steal your data and make money,” Fokker told The Register. “They’re not mythical. They don’t have superpowers.” And just like any other individual – or superhero – they sometimes slip up, and give the rest of us a moment of snarky joy. ®

Samsung’s next foldable could finally tackle three of the biggest complaints people have had about book-style folding phones for years.

A fresh leak has revealed what appears to be the most detailed look yet at the rumoured Galaxy Z Fold 8 Wide, which Samsung is expected to unveil alongside the Galaxy Z Fold 8 Ultra later this summer.

According to the leak, the Galaxy Z Fold 8 Wide could weigh just 201g, making it noticeably lighter than the current Galaxy Z Fold 7. That would put Samsung’s large-screen foldable surprisingly close to traditional flagship phones, while offering a a tablet-sized display when unfolded.

The leak also points to a 4.5mm thickness when open, alongside a larger 4,800mAh battery and support for 45W charging. If accurate, that would represent a welcome upgrade over previous Fold models which have often lagged behind rivals when it comes to battery capacity.

President Trump signed an executive order on Tuesday establishing a voluntary framework for government review of frontier AI models before public release, ending weeks of internal White House conflict over how aggressively to regulate the technology. The order, titled “Promoting Advanced Artificial Intelligence Innovation and Security,” was signed privately without the usual livestream or public ceremony, a contrast with the fanfare that typically accompanies presidential AI announcements.

The final version is substantially narrower than the draft Trump rejected on 21 May, when he scrapped a planned signing ceremony over concerns that the order “could dull America’s edge on AI technology.” The original draft proposed a 90-day mandatory pre-release review period and would have given the government formal evaluation authority over frontier models. The signed version asks companies to voluntarily submit models 30 days before release and participate in a collaborative framework rather than submitting to mandatory testing.

What the order does

The executive order establishes three main mechanisms. First, a voluntary pre-release review framework in which AI developers can engage the government to determine whether models under development qualify as “covered frontier models,” provide access for up to 30 days before planned release, and collaborate on selecting “trusted partners” for early access. The framework is explicitly voluntary, meaning companies can decline to participate without penalty.

Second, the order creates an AI cybersecurity clearinghouse within 30 days, coordinated by the Treasury Secretary, the National Cyber Director, the NSA, and CISA. The clearinghouse will scan for software vulnerabilities, validate discoveries, and coordinate remediation and patch distribution, a direct response to the Mythos crisis that demonstrated how AI-discovered vulnerabilities can outpace existing disclosure and patching processes.

Third, federal agencies are directed to develop benchmarks for assessing AI models’ cybersecurity capabilities and to strengthen the government’s own security defences against AI-enabled threats. The order also addresses AI safety research, though the specific provisions are less prescriptive than what the original draft contained.

What was cut

The differences between the scrapped draft and the signed order reflect the victory of the pro-industry faction within the White House. The 90-day mandatory review was reduced to a 30-day voluntary window. The formal government evaluation authority was replaced with a collaborative framework. The reporting requirements for companies developing powerful models, which would have echoed provisions in Biden’s repealed AI executive order, were softened to avoid what industry allies characterised as regulatory overreach.

Silicon Valley’s objections to the original draft were decisive. AI companies argued that mandatory pre-release testing would slow American innovation, create a competitive disadvantage relative to Chinese firms facing no equivalent requirements, and establish a precedent for government gatekeeping of technology deployment. The signed order addresses those concerns by making participation voluntary and framing the government’s role as collaborative rather than regulatory.

The gap it leaves

The voluntary framework means the order’s effectiveness depends entirely on whether AI companies choose to participate. Companies already engaged in pre-release testing with CAISI, including Google, Microsoft, and xAI, may continue or expand that cooperation. Companies that view government review as commercially disadvantageous or that are racing to ship products can simply opt out.

The EU’s AI Act, entering full enforcement in August, provides a stark contrast: mandatory requirements, statutory authority, and penalties for non-compliance. The Trump order establishes norms and creates institutional infrastructure (the cybersecurity clearinghouse, the benchmark development process) but relies on goodwill rather than obligation.

For the White House, the quiet signing may be the point. The order gives the administration a policy document it can reference when asked about AI oversight, creates structures that could be strengthened later, and avoids a public confrontation with an AI industry whose leaders are among the administration’s most visible supporters. Whether a voluntary framework is adequate for a technology that can discover 10,000 zero-day vulnerabilities in a month is the question the order deliberately leaves unanswered.