Tech

WhatsApp is rolling out a real-time voice chat feature for Meta AI on iOS beta version 25.21.10.76. This follows a previous rollout for Android users which we previously covered. When available, users can start a voice session by tapping on the waveform icon from the Chats tab, or by default, from the Calls tab. By rolling this out, Meta AI will catch up in a significant way with other chatbots which mostly offer calls already.

The integration on iOS has a few subtle differences to Android at the operating system level. When you begin a call with Meta AI, iOS will add an orange dot in the top-right corner of the screen to indicate that the microphone is in use. This privacy feature cannot be disabled by WhatsApp, so you’ll always know when Meta is listening.

Despite the rapid pace of GPU evolution and the hype around AI hardware, Linus Torvalds — the father of Linux — is still using a 2017-era AMD Radeon RX 580 as his main desktop GPU here in 2025. The Polaris-based graphics may be almost a decade old, but it’s aged remarkably well in Linux circles thanks to robust and mature open-source driver support. Torvalds' continued use of the RX 580, therefore, isn’t just boomer nostalgia. It's a statement of practicality, long-term support, and his disdain for unnecessary complexity.

Spotted by Phoronix, this revelation came during a bug report around AMD’s Display Stream Compression (DSC), which was causing black screen issues in Linux 6.17. Torvalds bisected the regression himself, eventually reverting a patch to maintain kernel progress. Ironically, DSC is what allows his Radeon RX 580 to comfortably drive his modern 5K ASUS ProArt monitor, a testament to how far open-source drivers have come.

“... same old boring Radeon RX 580,” Torvalds wrote in an email to the Linux Kernel Mailing List (LKML), reverting the patch for now so development can continue uninterrupted. That one line from the man himself speaks volumes about his preference for stability over novelty.

• Android Auto still doesn’t support the official YouTube app, but NewPipe can now fill that gap.
• The update lets you access playlists, history, and search from your car’s screen.
• Of course, you’ll only be able to watch while parked.

In a blog post, the NewPipe team says you can now browse your watch history, load playlists, or search for new content within Android Auto. The developers still urge caution when using the feature, even though Google already has guardrails to prevent you from getting distracted as you’re cruising along.

Amazon CEO Andy Jassy sees an opportunity to deliver ads to users during their conversations with the company’s AI-powered digital assistant, Alexa+, he said during Amazon’s second-quarter earnings call Thursday.

“People are excited about the devices that they can buy from us that has Alexa+ enabled in it. People do a lot of shopping [with Alexa+]; it’s a delightful shopping experience that will keep getting better,” said Jassy on the call with investors and Wall Street analysts. “I think over time, there will be opportunities, as people are engaging in more multi-turn conversations, to have advertising play a role to help people find discovery, and also as a lever to drive revenue.”

Amazon says it has rolled out Alexa+ to millions of customers, part of an effort to make its legacy digital assistant capable of agentic behaviors and more natural to talk to. Alexa+ is Amazon’s answer to generative AI voice assistants from OpenAI, Google, and Perplexity that have made legacy systems feel outdated. However, the business models behind generative AI products remain unclear.

Social media companies ​​are blocking wide-ranging content - including posts about the wars in Ukraine and Gaza - in an attempt to comply with the UK's new Online Safety Act, BBC Verify has found.

The new legislation, which came into effect last Friday, imposes fines on social media companies and other websites which fail to protect under-18s from pornography, posts promoting self-harm, and other harmful content. In serious cases, services could be blocked in the UK.

But BBC Verify found a range of public interest content, including parliamentary debates on grooming gangs, has been restricted on X and Reddit for those who have not completed age verification checks.

Experts warn companies are risking stifling legitimate public debate by overapplying the law.

Meta's addition of AI services to encrypted messaging platform WhatsApp has Italian officials suspecting the Silicon Valley giant may be abusing its dominant market position to push unwanted features on users.

The Italian Competition Authority (AGCM) said in a statement that Meta added its Meta AI service to WhatsApp "without any prior request from users" and that it was "placed in a prominent position on the screen and integrated into the search bar."

The AGCM believes these actions could be a violation of article 102 of the Treaty on the Functioning of the European Union (TFEU).

"By combining Meta AI with WhatsApp, Meta appears capable of channeling its customer base into the emerging market, not through merit-based competition, but by 'imposing' the availability of the two distinct services upon users, potentially harming competitors," the AGCM said.

That imposition locks users into Meta's AI ecosystem, the AGCM added, further entrenching Meta's market dominance and limiting competition from local AI firms. Meta has not directly disputed the possibility that integrating its AI into WhatsApp could confer market dominance.

The Competition and Markets Authority (CMA) convened a panel to investigate the cloud services market in the UK. The panel has now recommended that the CMA use its digital markets powers to prioritize an investigation into the largest cloud providers, Microsoft and Amazon Web Services, regarding their practices.

The panel's final decision highlighted three main issues. First, the UK cloud market is dominated by AWS and Microsoft, creating high barriers for new entrants to enter or grow. This situation gives Microsoft and Amazon significant market power and leaves customers with limited choice.

Second, technical and commercial barriers make it difficult for customers to switch cloud providers. This "lock-in" prevents them from taking advantage of better offers or more innovative services, which in turn weakens competition.

Third, Microsoft's software licensing strategies are specifically highlighted as anti-competitive. They are designed to disadvantage rivals like AWS and Google, further restricting choice for customers who rely on Microsoft software.

Apple and Google hold an “effective duopoly” with their mobile platforms and may be forced to open them up to more competition, a UK watchdog has said.

The Competition and Markets Authority (CMA) has announced proposals to give the US tech giants “strategic market status” for their mobile platforms, which could enforce changes that will benefit consumers, businesses and app developers.

An earlier market study by the CMA published in 2021 found that Apple and Google dominated mobile ecosystems across operating systems, app stores and web browsers.

It said this meant the two companies were in a position to effectively set the rules on how mobile browsers worked on their devices.

Around 90% to 100% of UK mobile devices run on Apple or Google’s mobile platforms, according to the CMA.

Sarah Cardell, chief executive of the regulator, said: “Apple and Google’s mobile platforms are both critical to the UK economy – playing an important role in all our lives, from banking and shopping to entertainment and education.

Microsoft has admitted that it can't protect EU data from U.S. snooping.

In sworn testimony before a French Senate inquiry into the role of public procurement in promoting digital sovereignty, Anton Carniaux, Microsoft France's director of public and legal affairs, was asked whether he could guarantee that French citizen data would never be transmitted to U.S. authorities without explicit French authorization. And, he replied, "No, I cannot guarantee it."

From Tom Scott's newsletter:

A lovely video from the design rules company: they launched a Kickstarter project, "Metroboard", and they visited Shenzen for a week to see the factories where it's being made. This video is ten minutes long, and the narration is calm and gentle, but the whole thing feels like it's going at breakneck speed: there is so much involved, and so much craft being shown. Thanks to Sebastian for sending this over!

Meta is hosting ads on Facebook, Instagram and Threads from pro-Israel entities that are raising money for military equipment including drones and tactical gear for Israeli Defense Force battalions, seemingly a violation of the company’s stated advertising policies, new research shows.

“We are the sniper team of Unit Shaked, stationed in Gaza, and we urgently need shooting tripods to complete our mission in Jabalia,” one ad on Facebook read, first published on 11 June and still active on 17 July.

The Wikimedia Foundation, which runs Wikipedia, is due to challenge the UK’s Online Safety Act (OSA) in the High Court of Justice in London on July 22 and 23. It wants to challenge the categorization regulations that would classify Wikipedia as a Category 1 service, which was designed for large, commercial social media platforms in mind, not volunteer, non-profit encyclopedias.

Around the beginning of last year, Matthew Prince started receiving worried calls from the bosses of big media companies. They told Mr Prince, whose firm, Cloudflare, provides security infrastructure to about a fifth of the web, that they faced a grave new online threat. “I said, ‘What, is it the North Koreans?’,” he recalls. “And they said, ‘No. It’s AI’.”

Those executives had spotted the early signs of a trend that has since become clear: artificial intelligence is transforming the way that people navigate the web. As users pose their queries to chatbots rather than conventional search engines, they are given answers, rather than links to follow. The result is that “content” publishers, from news providers and online forums to reference sites such as Wikipedia, are seeing alarming drops in their traffic.

As AI changes how people browse, it is altering the economic bargain at the heart of the internet. Human traffic has long been monetised using online advertising; now that traffic is drying up. Content producers are urgently trying to find new ways to make AI companies pay them for information. If they cannot, the open web may evolve into something very different.

Archive : https://archive.ph/nhrYS

LibreOffice has been on the offensive lately, taking the time to call out Microsoft and its practices whenever it can. Now, it is at it again, accusing Microsoft of "intentionally" using "unnecessarily complex" file formats to achieve user lock-in with its Microsoft 365 (Office) documents.

For those who don't know, XML is a markup language that programs like Microsoft 365 and LibreOffice use to structure and define documents.

As LibreOffice puts it:

An XML schema comprises the structure, data types and rules of an XML document and is described in an XML Schema Definition (XSD) file. This tells the PC what to expect and checks that the data follows the rules. In theory, XML and XSD together form the basis of the concept of interoperability.

The two office suites take very different paths here. LibreOffice uses the OpenDocument Format (ODF), an open standard meant to be controlled by no single company. This format gives us files like .odt for text and .ods for spreadsheets.

Microsoft, on the other hand, created its own Office Open XML (OOXML) to support every feature in its own software, giving us the familiar .docx and .xlsx. What's interesting is that both formats are really just ZIP archives. The easiest way to verify this is to take a .docx file, rename it to .zip, and decompress it. This will show you the guts of a Microsoft 365 document.

As LibreOffice notes, XML is supposed to function as "a bridge," but Microsoft is weaponizing its own schema by making it so "complex that it becomes a barrier rather than a bridge." LibreOffice compares it to a railway system where the tracks are public, but one company's control system is so convoluted that no one else can build a compatible train, making it almost impossible for others to compete. Passengers don't realise they are being held hostage by these technical hurdles.

While some creators are happy to see the growing capabilities of generative AI products like ChatGPT and Gemini, others are opposed to using AI tech, afraid that the AI could replace them. The movie industry is the best example of that. There’s concern that AI might take over jobs or alter performances, especially now that it’s more sophisticated than ever.

Products like Google’s Veo 3 can create lifelike video sequences that include dialogue and sound. With enough money and time, you could use such tools to make a full movie without hiring real actors or a production team.

Veo 3 isn’t the only advanced AI video generation tool out there, but it’s a good example of what’s possible today. And Google’s AI tech has already been used in at least one movie. Google said so a few months ago.

Fast-forward to mid-July, and Netflix confirmed that it used unnamed generative AI tools to create special effects for a widely acclaimed new TV show that became a monster hit earlier this year. The revelation came during Netflix’s quarterly earnings report, where the streamer reported a 16% rise in revenue for the June quarter.

Back in 1999, Wall Street lost its collective mind over the internet. Companies with no revenue were suddenly worth billions, “eyeballs” were treated as currency, and market analysts predicted a frictionless future where everything would be digital. Then the bubble burst. Between March 2000 and October 2002, an estimated five trillion dollars in market value vanished into thin air.

Today, it is happening again. This time, the magic word is not “.com.” It is “AI.” < According to Torsten Slok, the influential chief economist at Apollo Global Management, a major global investment firm, the current AI driven market bubble is even more stretched than the dot com frenzy of the late 1990s. And he has the data to prove it.

“The difference between the IT bubble in the 1990s and the AI bubble today is that the top 10 companies in the S&P 500 today are more overvalued than they were in the 1990s,” Slok wrote in a recent research note that was widely shared across social media and financial circles.

Supporting Windows 10 PCs is about to get very expensive, but upgrading to Windows 11 poses significant hardware and software challenges for many organizations. What terror awaits when 1,000 applets IT never knew about crash after the upgrade?

For IT leaders, there is much to fear. How much of their current hardware is compatible with Windows 11? Far more critically, what about operational technology devices that manage industrial processes, on-prem legacy apps (including tons of homegrown code), and the unknown numbers of applets that IT doesn’t know about?

That “unknown” list includes inherited apps from the company’s last 50 acquisitions, as well as shadow apps that business units never bothered to report to corporate.

At the same time, there is a noticeable lack of enthusiasm among IT leaders for making the upgrade at all, given the perception that Windows 11 simply doesn’t offer much in terms of materially new or better functionality.

The decision is being forced, because Microsoft says that it will not add new capabilities or provide security patches for Windows 10 for corporate customers after October 14, 2025 — unless they enroll in the Extended Security Updates (ESU) program for Windows 10.

Microsoft is doubling the ESU price every year: For the first year, it will be $61/device, which will rise to $122/device for the second year and then $244/device for the third year. After that, the company says it will cut off all Windows 10 support entirely. (Microsoft did not respond to a Computerworld request for an interview for this story.)

Enrolling 5,000 Windows 10 PCs in ESU for the full three years would cost a business more than $2.1 million. A large organization that wants to keep 30,000 PCs on extended Windows 10 support for three years would have to pay more than $12.8 million to do so.

This leaves IT leaders having to choose between a potentially painful upgrade to Windows 11 and having to pay a massive amount of money for continued Windows 10 support, depending on how many seats will remain on Windows 10.

WeTransfer was forced to respond this week after changes to its terms of service (TOS) triggered major backlash from users who believed the new language granted the service access to users' files to train AI.

"We don't use machine learning or any form of AI to process content shared via WeTransfer, nor do we sell content or data to any third parties," a WeTransfer spokesperson told BBC News on Tuesday.

WeTransfer clarified this after users noticed recent changes to its TOS page, which initially said the following policy would go into effect in August (via Wayback Machine on July 14, 2025).