In brief: Aside from concerns about human extinction or a massive financial bubble, one of generative AI’s biggest controversies is how tech giants train their models. While OpenAI and Microsoft claim that training AI on copyrighted material falls under fair use, internal comments from the companies could suggest otherwise.

Brent Hecht, Microsoft’s director of applied science, predicted that people worldwide would view large language models as theft on an unprecedented scale, according to court documents released Thursday. The comments appear to counter Microsoft and OpenAI’s public defense against an ongoing lawsuit brought by news outlets.

In the case’s latest filing, Hecht said generative AI could be the largest labor theft in human history while describing how Microsoft, OpenAI, and other AI companies train their models on massive amounts of material from the open internet. The practice lies at the heart of the copyright lawsuit that The New York Times filed in late 2023.

Microsoft and OpenAI argue that training AI models on articles from the NYT and other publications falls under fair use, similar to a student reading books. The defendants also claim that generative AI output based on those articles sufficiently transforms the original material.

However, Thursday’s filing contains internal comments from OpenAI that highlight AI’s ability to reproduce news articles. For example, OpenAI executive Greg Brockman admitted that ChatGPT can predict and complete sentences from NYT articles when working with the outlet’s material. Hecht also argued that winning the lawsuit might require the defendants to “make a complete mockery of the idea of fair use.”

OpenAI has also previously admitted that training AI without accessing copyrighted material is impossible. Former Meta executive Nick Clegg echoed that view last year, claiming AI would die almost immediately if copyright law were enforced against it.

Meanwhile, AI companies have continued training their models on reams of user data, articles, and other material. Meta drew controversy for training its AI on employee behavior, Microsoft’s GitHub Copilot trains on user data unless users opt out, and Twitch only started allowing streamers to opt out after training on their material for years. AI developers are also quietly buying, scanning, and destroying millions of books to gather more training material.

Generative AI’s impact on the news business model is another pillar of the NYT’s lawsuit. Google and other search engines now provide AI-generated summaries of information from articles in response to user queries, potentially discouraging readers from visiting news websites. In this week’s court filing, an OpenAI executive involved in ChatGPT’s development said publishers face an existential threat from AI.

Leave it to IKEA to name its new indoor Bluetooth speaker BADKRUKA, a Swedish word for someone who is reluctant or afraid to get into cold water. That is particularly appropriate because the $59.99 BADKRUKA carries only an IP20 rating and has absolutely no business going swimming.

The name may be amusing, but IKEA has taken a surprisingly practical approach to the speaker itself. BADKRUKA combines Bluetooth with 3.5mm analog and optical digital inputs, along with Spotify Tap, three EQ settings, multi-speaker operation and the ability to pair two BADKRUKA speakers for actual stereo playback. IKEA even includes the optical cable.

That does not suddenly make BADKRUKA a $60 replacement for a serious pair of active loudspeakers. We have not heard it yet. But IKEA has given budget buyers considerably more flexibility than we normally see at this price.

IKEA has been rebuilding its audio lineup since ending its long-running partnership with Sonos. We previously looked at the company’s BLOMPRAKT and NATTBAD Bluetooth speakers and the end of the IKEA and Sonos SYMFONISK collaboration.

More Than Another Bluetooth Box

BADKRUKA uses a woofer and tweeter configuration powered by 25 watts of total amplification, with IKEA specifying a frequency range of 50Hz to 20kHz at -6 dB. The enclosure measures 6.25 inches wide and deep and 6.875 inches tall, making it small enough for a desk, bedroom or shelf beside a television.

The optical input is what changes the equation.

Most modern televisions still provide an optical digital output, which means BADKRUKA can function as a very inexpensive alternative to the television’s built-in speakers. Buy two and IKEA says they can operate as a stereo pair with distinct left and right channels.

There is no HDMI ARC here, so television volume control will not be as seamless as it would be with a soundbar or more sophisticated active loudspeaker system. IKEA also does not list Wi-Fi streaming, AirPlay, Google Cast or other network playback features.

At $59.99, that feels less like an omission than a reminder of what this thing costs.

Spotify Tap and Multi-Speaker Playback

BADKRUKA supports Spotify Tap, allowing playback to resume from a paired device with a button press. Pressing the Spotify button again can generate another selection based on the listener’s history.

IKEA has also included its multi-speaker mode, which allows BADKRUKA to play alongside other compatible IKEA Bluetooth speakers. That mode distributes the same mono signal to multiple speakers, while stereo pairing is reserved for two BADKRUKA units.

Three built-in EQ settings provide some control over the presentation, although IKEA does not provide detailed response curves for the presets.

BADKRUKA vs. NATTBAD

The more interesting comparison comes from inside IKEA’s own catalog.

The existing NATTBAD costs $49.99 and provides 10 watts of amplification through a single 2.8-inch full-range driver, with a specified frequency range of 75Hz to 18.5kHz. It supports Bluetooth, Spotify Tap and multi-speaker playback but does not offer stereo pairing, optical or 3.5mm analog input.

For another $10, BADKRUKA increases rated output to 25 watts, uses separate woofer and tweeter drivers, extends the claimed low-frequency response to 50Hz and adds both wired inputs and stereo operation.

We have not compared the two sonically, so specifications should not be confused with a listening test. But unless the smaller NATTBAD design is the primary attraction, BADKRUKA looks like the much stronger value on paper.

IKEA BADKRUKA Specifications

• Speaker Configuration: Woofer and tweeter
• Total Power Output: 25 watts
• Frequency Range: 50Hz to 20kHz (-6 dB)
• Wireless Input: Bluetooth
• Digital Input: Optical TOSLINK
• Analog Input: 3.5mm
• Spotify Tap: Yes
• Stereo Pairing: Yes
• Multi-Speaker Mode: Yes
• EQ: 3 settings
• IP Rating: IP20
• Dimensions: 6.25 x 6.25 x 6.875 inches
• Included Accessory: Optical cable
• Designer: Ola Wihlborg

The Bottom Line

We are not expecting the IKEA BADKRUKA to send KEF, Sonos or Audioengine executives scrambling into emergency meetings. That would be asking rather a lot from something sold by the same company that expects you to assemble your own dresser with an Allen key.

What makes BADKRUKA interesting is that IKEA has resisted turning it into another Bluetooth-only lifestyle product. Optical and analog inputs make it useful with televisions, computers and other sources, while stereo pairing gives buyers an inexpensive upgrade path if one speaker is not enough.

At $59.99 each, a pair costs $119.98.

That might be the more interesting story.

Price & Availability

The IKEA BADKRUKA Bluetooth Speaker is $59.99 each at IKEA in the United States in black and pink finishes. Availability may vary by store and delivery location.

Related Reading:

Uh oh, we got a cute alert! It’s Bezzy, your adorable, cuddly, anthropomorphized server-rack and stuffed-animal friend! Give this soft lil’ doll a squeeze, and it shall sing you its telltale song: the piercing, grinding ambient squeals emitted from a real US data center.

The Bezzy doll is a stuffed plushie made in a satirical collaboration between the New York–based creative studio Basura and Big Data, the musical project of producer Alan Wilkis, best known for the song “Dangerous,” a 2013 collaboration with the band Joywave. The joint project, which is coming out the same day as a new Big Data single, is called Adopt a Data Center. The goal is to present people with a “cuddly introduction to life with a real data center.”

Data Centers are very un-vogue right now. Built out by all the big AI companies to power their increasingly power-hungry AI tools, data centers have spread across thousands of acres of land in the US. They have been heavily criticized for their energy consumption and expansion of gas power. They have also proven to be audibly annoying, stirring up complaints about noise pollution from those nearby.

The Adopt a Data Center stunt aims to make people think differently about how services produced by data centers fit into their lives. That manifests as a plush stuffie that looks like a server rack in a data center, albeit with long arms and legs and googly eyes. It’s cute in the way Don’t Hug Me I’m Scared is cute. Squeeze Bezzy’s torso and the sound bursts out. It’s a screeching few seconds of a recording taken by a neighbor of a data center in Virginia. The tag on the Bezzy plushie reads, “Hug me for ear-piercing forever hum.”

“We picked this recording as it was the clearest and most suitably irritating,” wrote Rajeev Basu, the founder of Basura, in an email to WIRED. In an interview with the person who recorded the audio, Basu says they described the sound as “like a car alarm going off nonstop. It’s high-pitched, and it’s all you can focus on.”

Basura has worked on other projects like making luxury wear for dogs and a partnership with Yerba Madre that made dirt shoes designed to break apart during a walk to spread wildflower seeds.

Along with Bezzy, the data centers project includes a collection of stickers that say things like “Every data center deserves a forever home” and “Open your heart. And your power grid.” Basu says the project is the result of dozens of interviews with people who live near data centers in the US. The website itself reads, “Adopt a data center: All they ask for is everything.”

Highly anticipated: We’re only a few weeks away from the release of Gears of War: E-Day, and that means the final PC system requirements have been revealed. This is yet another game where all the listed configurations use upscaling, so hitting 4K/60+ FPS with an RTX 5070 Ti isn’t achieved natively.

Gears of War: E-Day’s system specs first appeared on its Steam page back in June, but the listings were for Minimum and Recommended specs, with no mention of resolutions, presets, upscaling, etc.

Now, developer The Coalition has revealed the final set of requirements ahead of the game’s October 6 release.

The minimum specs remain the same, requiring just an RTX 5050/RTX 2060 or Radeon RX 6600/RX 9060 combined with an Intel i7-6850K/i5-10400/Ryzen 5 2600X. We now know that this uses a Medium graphics preset and aims for 60 FPS at 1080p.

However, we also know that all the listings are using upscaling, though there’s no mention of whether they are set to Quality, Balanced, Performance, or Ultra Performance. The recommended configuration uses a High graphics preset and aims for 1440p at 60 FPS and higher.

Interestingly, The Coalition has removed the Arc A580 (minimum) and Arc B580 (recommended) from the hardware list. It does list Intel XeSS 3.0, so those with Intel GPUs shouldn’t worry.

The new addition is the Ultra spec category: 4K/60+ FPS at Ultra graphics. It asks for at least an RTX 5070 Ti or Radeon RX 7900 XT combined with an Intel i7-14700K or Ryzen 7 7800X3D.

Another change is the storage requirement. It previously asked for 130GB of free SSD space, but that’s been reduced by 15GB to 115GB. It’s still a lot, but any reduction is welcome.

There’s also a disclaimer at the bottom that warns graphics cards without dedicated hardware ray tracing support are not supported by the next Gears of War entry.

It seems that most games’ system requirements these days include the use of upscaling. It used to be that studios would be less forthcoming about this fact, but this is changing. Control Resonant, for example, even shows the actual upscaler preset for each requirement level. The PC requirements for Blood of the Dawnwalker also specified the use of upscaling, but the recently revealed hardware requirements for Fable don’t clarify whether the performance targets are achieved natively.

Pre-installation for Gears of War: E-Day begins on September 29. Premium Edition owners can start playing on October 1, five days before the full launch.

Like the Galaxy S25 FE, the S26 FE feels more like the Galaxy S26+ than the S26, with a larger 6.7-inch display and a trio of cameras on the back. It’s hard to spot the differences from its predecessor since there are only a few minor design changes, including a new glossy back finish. It’s a choice: The S25 FE had a matte finish, as does the flagship S26 family. I guess it’s been done to differentiate, but I’m not sure why. The newest Galaxy phone is also marginally taller, wider and heavier (3 grams) than the S25 FE, though the 7.4mm thickness remains identical. It also has a slightly more pronounced chin at the bottom of the device.

Otherwise, it looks and feels like a flagship Galaxy phone that costs several hundred dollars more. Plus, I love the pistachio colored device I received for review. I’m always happy to see more diverse color options appear alongside the monochrome status quo.

The S26 FE also has a new connected camera island, replacing the camera lenses that slightly jutted out of the back of the S25 FE. It’s a minor design tweak, but it does help unify the device with both the Galaxy S and Z families.

The 6.7-inch OLED display looks crisp and vivid, though anyone coming from flagship devices will notice the lower max brightness on sunny summer days. It peaks at 1,900 nits, while the Galaxy S26 series cranks out up to 2,600 nits. The display does, however, reach up to 120Hz, so it’s at least a smooth experience, and compared to the base S26, I appreciate having more screen on a device that’s currently $200 less.

Diving into the rest of the technical specs reveals the tough economic choices made to keep the price low. That said, it’s still $50 more than the S25 FE, with the same 8GB of RAM and only 128GB of storage on the base iteration. This fills up quickly, especially with asset-dense games like Wuthering Waves. (There is a 256GB version, if needed.)

This year’s Fan Edition now has an Exynos 2500 chip, which makes it at least a little more capable for gaming. It’s certainly not a flagship experience, however. The S26 FE struggled with the demands of Wuthering Waves, with frame rates often choking in prettier, more detailed environments or hectic fights. Some may argue this is one sign that this isn’t a fully fledged Galaxy S phone, but it uses the same chip as last year’s Z Flip 7. Unless you’re pushing the phone to its limits, it’s otherwise fast and responsive most of the time.

One thing I did note was that the camera app was occasionally so sluggish it felt like a much cheaper A-series phone. This was intermittent, but also frustrating. Trying to snap an in-the-moment shot demands a zippy camera app. It’s been years since phone companies have bragged about how responsive their camera launchers and shutter are, because they are all pretty swift now, making the S26 FE’s struggles all the more jarring.

At least the battery’s big, matching the 4,900mAh cell in the similarly sized S26+ (and last year’s FE phone, again). In our video rundown test, we eked out just over 21 hours of playback in day-to-day use. Despite the larger battery than what’s in the S26, that’s a big drop in longevity. This means the S26 FE demanded daily charging. Those speeds top out at 45-watt wired charging with support for 15W wireless charging, too. No magnetic rings for mounting, but hey, at least there’s wireless charging.

Microsoft has resolved a known issue that causes incorrect alerts warning that Defender Antivirus was turned off after installing recent updates.

In a Windows release health dashboard update on Thursday, Microsoft said the issue was fixed in the Microsoft Defender Antivirus update (version 4.18.26080.4) released on September 17.

The company acknowledged the bug in late August, even though the issue had affected users in the Release Preview Channel of the Windows Insider program since at least June.

As Microsoft explained, this affects all supported Windows client and server versions, including the latest Windows 11 26H1 and Windows Server 2025 releases, and it triggers erroneous alerts in the Windows Security app that prompt users to “Tap or click to turn on Microsoft Defender Antivirus.”

“After installing the latest updates for Microsoft Defender Antivirus, notifications might appear stating that “Microsoft Defender Antivirus is turned off,” even though the antivirus is functioning correctly and all settings show it as active,” it said at the time.

“These notifications can appear when Windows starts and intermittently afterward. They persist even if notification settings are turned off.”

​This isn’t the first time Microsoft asked customers to ignore incorrect errors and alerts displayed on their systems after installing updates.

In April 2025, the company addressed an issue that was triggering incorrect BitLocker drive encryption errors on Windows 10 and Windows 11 devices and fixed a bug that caused invalid 0x80070643 failure errors after installing the Windows Recovery Environment (WinRE) updates.

It also asked users in July 2025 to ignore erroneous Windows Firewall alerts that appeared after rebooting after installing the June 2025 preview update.

One month later, it warned that the July 2025 preview update and subsequent Windows 11 24H2 updates were causing incorrect CertificateServicesClient (CertEnroll) errors.

This week, Microsoft also released emergency Windows updates to fix Remote Desktop Services, Hyper-V, and USB audio issues caused by the September 2026 security updates.

Join Mikko Hyppönen and security leaders from the NFL, CHANEL, and Atlassian for a two-hour digital summit on what AI-speed attacks change, what defenders should stop doing, and how to validate, decide, fix, and re-validate at machine speed.

Save your seat

Tesla Charging posted the first Accordion Supercharger on September 15 and laid the plan out in one short clip. Factory work now covers the posts, the cabinet, the wiring, the cooling lines, and the concrete pads so a finished island leaves the plant instead of a pile of parts. Two of those eight-stall packages fit on a single truck, which means sixteen V4 stalls arrive in one run. Tesla says that package costs about 20 percent less to put in the ground than a conventional station and commissions as soon as utility power is tied in.

Factory crews place each pair of stalls on a prefabricated platform and simply insert a central V4 cabinet rated at 1.2 megawatts. Heavy gage conduits extend from the cabinet to the pedestals, allowing the island to fit across a painted parking bay. That’s crucial, because there won’t be enough curb strip on retail properties to dig a standard trench for a cabinet. To set up, a forklift or telehandler takes the entire bundle off the trailer, personnel lower it into a shallow trench, the pedestals spread outwards, the trench is filled back in, pavers are placed in, and the utility feed is connected. There is no need to complete any DC busbar work on-site. There’s no need to call a Tesla service professional to switch everything on either.

Sale

BLUETTI Elite 30 V2 Portable Power Station 288Wh 600W LFP Solar Generator

• [288Wh On-the-Go Power] – Only 9.4 lbs lightweight, carry it anywhere during storms! 288Wh capacity meets daily & outdoor needs—camping, road trips…
• [600W Continuous & 1500W Surge Power] – Get a full 600W output—twice as much as others. When you need serious power, activate Power Lifting Mode for…
• [50% Lower Power Consumption] – Upgraded UltraCell tech & smart cooling system cut power consumption by 50%. Enhances overall energy efficiency…

The first real example is already in Irschenberg, Bavaria, and it includes eight V4 Superchargers on display at the FC Bayern store lot, sandwiched between two fast food restaurants. The site became active in August, and each Supercharger can deliver 500 kilowatts to a car. The total combined cabinet power is 1.2 megawatts, which already exceeds the 1,050 kilowatt minimum that Europe will require from all charging pools on major routes by the end of 2027, and they must be located within 60 kilometers of one another.

This variation was inspired by the variety of parking lots found throughout Europe. Tesla unveiled the Folding Unit (FU) a few months ago, with the idea that two fixed portions would fold out along a curb, reducing manufacturing time from eight to four days. Accordion maintains the core idea of being manufactured in a factory first, but with a twist: the pedestals now stretch as needed between rows rather than attempting to line up behind a curb. That is how most supermarket and shop lots are painted. Max de Zegher, the man in charge of Tesla Supercharging, put it simply: building in a factory is preferable to building on-site, and all jobs and servicing comments go directly back to the production line, so the following unit is already better.

As a result, the parking lot owners receive a finished product without having to deal with weeks of digging trenches, laying conduits, erecting pads, and hammering in single posts. According to Tesla, the plan is to pre-assemble each Supercharger in a factory, with the Accordion version reserved for sites that will not accommodate a cabinet behind the curb. You get 16 stalls per shipment, 20% less installation cost, rapid commissioning, and V4 hardware that can already meet the European power norm for 2027, with the first one now operational in Bavaria; the next ones will demonstrate whether this manufacturing concept can keep up with trucks.

Sounding off: We hear a lot about US views of AI and its effect on jobs, but does the rest of the world agree? Yes, according to a new global Pew Research Center poll, which found that more people in 34 out of 37 countries believe AI will lead to fewer rather than more jobs.

The US was the third most pessimistic nation when it comes to AI’s effect on the workplace, with 71% of participants saying it will lead to fewer jobs in the country over the next 20 years. Only South Korea and Australia, both on 76%, were higher.

Across a 37-country median, 46% said there will be fewer jobs, 13% said there wouldn’t be much difference, 9% said there will be more jobs, and 25% weren’t sure.

The table illustrates the disparity between nations with higher and lower GDPs. The “fewer jobs” result is much lower in poorer countries: Nigeria was the only one to believe AI will create more jobs (26% vs. 27%), while more people in Thailand and the Philippines picked the “not sure” option. It’s likely that this is a reflection of fears in wealthier regions that AI will take mostly white-collar jobs – as Dario Amodei predicted.

Interestingly, the only country where Pew has asked this question before is the US. Over the past two years, the share of people who believe more jobs will be lost from AI has increased 7 percentage points, showing that fears are intensifying rather than easing in the States.

In a number of countries, younger people aged between 18 and 34 are the group most concerned about AI job losses. In the US, concern about job losses from AI has increased the most among this demographic.

In many of the middle-income countries surveyed, wealthier and more educated respondents are more likely to expect AI to reduce job numbers over the next two decades. Much of that gap reflects greater uncertainty among people with lower incomes and less education about how the technology will affect employment.

It’s certainly telling that more people who have high awareness of AI are more concerned about its impact on jobs.

Some industry execs have claimed AI will close the gap between the rich and the poor. But more people believe the opposite is true. Again, this is more often the case in richer countries.

We’ve also heard of some AI luminaries who are genuinely angry that people don’t share their excitement about AI. Results in this area are more mixed, with a 37% median across 37 countries more concerned than excited, while a 41% median are equally concerned and excited. Israel is the only nation in which more people are primarily excited than primarily concerned.

Those fears are understandable. A Mercer survey found that 99% of executives questioned expected AI to lead to job cuts within two years, while Meta cut 10% of its workforce before abandoning plans for a second wave amid disappointing AI results and employee resistance. There have also been signs that companies are putting too much faith in the technology: some employers are rehiring workers they replaced with AI after automation failed to deliver. That offers a bit of hope, though it’s unlikely to reassure all workers.