verge-rss

Cath Virginia / The Verge

An antitrust investigation by the Department of Justice focusing on Nvidia’s AI dominance as the hardware company of choice is escalating, reports Bloomberg. The outlet reports that Nvidia and other companies have received legally binding requests for information as its sources say regulators are investigating whether Nvidia is “making it harder to switch to other suppliers and penalizes buyers that don’t exclusively use its artificial intelligence chips.”

Earlier this summer, Nvidia had overtaken Microsoft as the world’s most valuable company, but now it trails both Microsoft and Apple. Before the subpoenas were reported on Tuesday afternoon, its share price dropped more than 9 percent, wiping out about $279 billion in market value in one day.
But as Bloomberg notes, Nvidia’s stock is still up more than double this year as the AI boom has fueled sales of the company’s pricey chips, even after a report by The Information a few weeks ago saying that its upcoming Blackwell series of AI chips may be delayed “by three months or more due to design flaws.”

Cath Virginia / The Verge

An antitrust investigation by the Department of Justice focusing on Nvidia’s AI dominance as the hardware company of choice is escalating, reports Bloomberg. The outlet reports that Nvidia and other companies have received legally binding requests for information as its sources say regulators are investigating whether Nvidia is “making it harder to switch to other suppliers and penalizes buyers that don’t exclusively use its artificial intelligence chips.”

Earlier this summer, Nvidia had overtaken Microsoft as the world’s most valuable company, but now it trails both Microsoft and Apple. Before the subpoenas were reported on Tuesday afternoon, its share price dropped more than 9 percent, wiping out about $279 billion in market value in one day.

But as Bloomberg notes, Nvidia’s stock is still up more than double this year as the AI boom has fueled sales of the company’s pricey chips, even after a report by The Information a few weeks ago saying that its upcoming Blackwell series of AI chips may be delayed “by three months or more due to design flaws.”

Read More 

Photo by Allison Johnson / The Verge

Last month, it was discovered that an inactive piece of software that ships on all Google Pixel smartphones presented a potential security vulnerability for device owners. The company quickly said it would remove “Showcase.apk,” which was created years ago “for Verizon in-store demo devices and is no longer being used.” But on the off chance it were ever exploited, cybersecurity firm iVerify found that the software could make phones “accessible to hackers and ripe for man-in-the-middle attacks, code injection, and spyware.” The flare-up was enough for Palantir to implement a companywide ban on Android devices.
At the time, Google told us that it had “seen no evidence of any active exploitation,” and with today’s release of the monthly Pixel software update, it’s getting rid of Showcase.apk altogether. The release notes confirm that the September patch includes a fix to “remove third-party APK to address security vulnerability.”
Other than that, it’s a very small update, with improvements to Wi-Fi stability and performance for Google’s brand-new Pixel 9 portfolio as the only other bullet point. Bigger software upgrades await next month; that’s when Google plans to roll out Android 15 to its devices. The major OS update was released to the Android Open Source Project (AOSP) today, but it’ll be a few more weeks until it’s ready for Pixel hardware.

Photo by Allison Johnson / The Verge

Last month, it was discovered that an inactive piece of software that ships on all Google Pixel smartphones presented a potential security vulnerability for device owners. The company quickly said it would remove “Showcase.apk,” which was created years ago “for Verizon in-store demo devices and is no longer being used.” But on the off chance it were ever exploited, cybersecurity firm iVerify found that the software could make phones “accessible to hackers and ripe for man-in-the-middle attacks, code injection, and spyware.” The flare-up was enough for Palantir to implement a companywide ban on Android devices.

At the time, Google told us that it had “seen no evidence of any active exploitation,” and with today’s release of the monthly Pixel software update, it’s getting rid of Showcase.apk altogether. The release notes confirm that the September patch includes a fix to “remove third-party APK to address security vulnerability.”

Other than that, it’s a very small update, with improvements to Wi-Fi stability and performance for Google’s brand-new Pixel 9 portfolio as the only other bullet point. Bigger software upgrades await next month; that’s when Google plans to roll out Android 15 to its devices. The major OS update was released to the Android Open Source Project (AOSP) today, but it’ll be a few more weeks until it’s ready for Pixel hardware.

Read More 

Castlevania: Dawn of Sorrow. | Image: Konami

For anyone looking to explore the best Castlevania has to offer, the just-launched Castlevania Dominus Collection is pretty much a no-brainer. It includes a trilogy of games from the Nintendo DS era that show designer Koji Igarashi arguably at the height of his powers: the experiences combine smooth, satisfying action with worlds that are intricate labyrinths to be mastered slowly over time. Bringing that experience to modern platforms alone makes the collection worth it. But the Dominus Collection also goes a step further and includes some smart options for porting over a dual-screen game. Then, it throws in an unexpectedly excellent remaster just for fun.
Part of what made Castlevania such a good fit for the DS was how much information could be displayed at once. While the side-scrolling action took place on one screen, the other could show character information or, more crucially, the map. For a game that’s all about traversing a complex castle, with lots of backtracking and blocked entrances to contend with, having that map always on is incredibly helpful. And not only is that option still present in this modern rerelease but Konami has also given players a few choices for how they want their info displayed.
By default, the Dominus Collection — which includes Dawn of Sorrow, Portrait of Ruin, and Order of Ecclesia and is available on PlayStation, Switch, Xbox, and PC — displays the games as shown in the screenshot above. The action is largest on the left side of the screen, while both the map and character info are displayed beside it. There are a few other options, though, which really let you customize the experience. For two examples: you can have just the action and the map side by side, or you can have it top and bottom to mimic an actual DS, like this:

I’ve been playing through Portrait of Ruin mostly using the side-by-side option, but the fact that there’s so much choice here is important. It’s a thoughtful approach to porting games from a distinct piece of hardware and means that these excellent titles are available in a way that both feels native to modern platforms and honors the original releases.
There’s also an added bonus included in the Dominus Collection. While Castlevania is a series with some beloved highs, it has also had some notable duds over the years. And while the arcade spinoff Haunted Mansion always looked good, it plays like, well, a game that’s trying to suck up all of your quarters. It’s hard and tedious and missing that Castlevania oomph. Including it in this collection is therefore a surprising choice but also one that is very welcome because, in addition to the stodgy original, you also get a full-on remake that spruces up the graphics and tweaks the design and difficulty to make it actually fun to play. It turns an inaccessible arcade release into a decidedly old-school Castlevania, from before the series shifted to an exploration focus.
It’s also the kind of oddity that’s perfect for a retro bundle like this. In the absence of a playable documentary, the Dominus Collection is about as good as it gets.

Castlevania: Dawn of Sorrow. | Image: Konami

For anyone looking to explore the best Castlevania has to offer, the just-launched Castlevania Dominus Collection is pretty much a no-brainer. It includes a trilogy of games from the Nintendo DS era that show designer Koji Igarashi arguably at the height of his powers: the experiences combine smooth, satisfying action with worlds that are intricate labyrinths to be mastered slowly over time. Bringing that experience to modern platforms alone makes the collection worth it. But the Dominus Collection also goes a step further and includes some smart options for porting over a dual-screen game. Then, it throws in an unexpectedly excellent remaster just for fun.

Part of what made Castlevania such a good fit for the DS was how much information could be displayed at once. While the side-scrolling action took place on one screen, the other could show character information or, more crucially, the map. For a game that’s all about traversing a complex castle, with lots of backtracking and blocked entrances to contend with, having that map always on is incredibly helpful. And not only is that option still present in this modern rerelease but Konami has also given players a few choices for how they want their info displayed.

By default, the Dominus Collection — which includes Dawn of Sorrow, Portrait of Ruin, and Order of Ecclesia and is available on PlayStation, Switch, Xbox, and PC — displays the games as shown in the screenshot above. The action is largest on the left side of the screen, while both the map and character info are displayed beside it. There are a few other options, though, which really let you customize the experience. For two examples: you can have just the action and the map side by side, or you can have it top and bottom to mimic an actual DS, like this:

I’ve been playing through Portrait of Ruin mostly using the side-by-side option, but the fact that there’s so much choice here is important. It’s a thoughtful approach to porting games from a distinct piece of hardware and means that these excellent titles are available in a way that both feels native to modern platforms and honors the original releases.

There’s also an added bonus included in the Dominus Collection. While Castlevania is a series with some beloved highs, it has also had some notable duds over the years. And while the arcade spinoff Haunted Mansion always looked good, it plays like, well, a game that’s trying to suck up all of your quarters. It’s hard and tedious and missing that Castlevania oomph. Including it in this collection is therefore a surprising choice but also one that is very welcome because, in addition to the stodgy original, you also get a full-on remake that spruces up the graphics and tweaks the design and difficulty to make it actually fun to play. It turns an inaccessible arcade release into a decidedly old-school Castlevania, from before the series shifted to an exploration focus.

It’s also the kind of oddity that’s perfect for a retro bundle like this. In the absence of a playable documentary, the Dominus Collection is about as good as it gets.

Read More 

Image: Volkswagen

Volkswagen has announced it’s rolling out its improved in-car AI voice assistant, IDA, with both locally processed and cloud-based responses from ChatGPT and other models, starting on September 6th. The first cars getting the feature will be the 2025 Jetta and Jetta GLI and the 82-kWh battery version of the 2024 ID.4, with other cars getting it later this year.
Drivers in cars that support ChatGPT will need to be subscribed to VW’s Plus Speech with AI service to get it. Down the line, the company says most of its 2025 models will get the new AI voice assistance features, including the ID.4 and ID Buzz (which will both get it free for three years), and the GTI, Golf R, and Tiguan (free for one year). Jetta, Jetta GLI, and Taos owners will need to subscribe on their own. Atlas and Atlas Cross Sport models won’t get ChatGPT until 2026, VW says.

The company had previously announced its ChatGPT integration plan at CES in January. At the time, it wasn’t clear how it all would work. VW says that OpenAI’s chatbot, along with a “multitude” of other models are provided by automotive chatbot company Cerence, will take over for IDA when requests are more complex than tweaking your air conditioning settings. For instance, the company says requests for things like restaurant suggestions or to be told a story will go to the cloud.
IDA can locally process natural language requests for certain things, VW claims. Tell it you’re “chilly,” and it should know you mean you want the heat on. (Here’s hoping it doesn’t sometimes kick that over to ChatGPT to tell you about nearby chili restaurants.)

Image: Volkswagen

Volkswagen has announced it’s rolling out its improved in-car AI voice assistant, IDA, with both locally processed and cloud-based responses from ChatGPT and other models, starting on September 6th. The first cars getting the feature will be the 2025 Jetta and Jetta GLI and the 82-kWh battery version of the 2024 ID.4, with other cars getting it later this year.

Drivers in cars that support ChatGPT will need to be subscribed to VW’s Plus Speech with AI service to get it. Down the line, the company says most of its 2025 models will get the new AI voice assistance features, including the ID.4 and ID Buzz (which will both get it free for three years), and the GTI, Golf R, and Tiguan (free for one year). Jetta, Jetta GLI, and Taos owners will need to subscribe on their own. Atlas and Atlas Cross Sport models won’t get ChatGPT until 2026, VW says.

The company had previously announced its ChatGPT integration plan at CES in January. At the time, it wasn’t clear how it all would work. VW says that OpenAI’s chatbot, along with a “multitude” of other models are provided by automotive chatbot company Cerence, will take over for IDA when requests are more complex than tweaking your air conditioning settings. For instance, the company says requests for things like restaurant suggestions or to be told a story will go to the cloud.

IDA can locally process natural language requests for certain things, VW claims. Tell it you’re “chilly,” and it should know you mean you want the heat on. (Here’s hoping it doesn’t sometimes kick that over to ChatGPT to tell you about nearby chili restaurants.)

Read More 

Image: Nick Barclay / The Verge

People are losing a lot more money to Bitcoin ATM scams. In a report released on Tuesday, the Federal Trade Commission said it found the amount of money lost to Bitcoin ATM scams increased nearly 10 times from 2020 to 2023 — going from $12 million to a whopping $114 million. Consumers have already lost $65 million to the ruse in the first half of 2024 alone.
Over the years, scams have evolved into different versions of pretty much the same thing: schemes that trick victims into paying scammers. We’ve seen bad actors trick people into sending wire transfers, buying gift cards, and even handing over a pile of cash in a shoebox. This is the Bitcoin ATM variation.
Bitcoin ATMs are the physical machines that you can use to purchase or sell cryptocurrency at places like gas stations and grocery stores. To carry out their scheme, scammers contact a victim by phone, text, or through online pop-ups. They’ll then impersonate a bank or government official while attempting to convince their victim that their bank account has been compromised and that they need to take action.

Image: FTC

Scammers will then instruct their victims to take out a large sum of cash and deposit it into a Bitcoin ATM — which the FTC says scammers may call “safety lockers” — to keep their funds secure. Once at the Bitcoin ATM, scammers will send their victim a QR code to scan at the machine, which will put all of the victim’s cash into the scammer’s crypto wallet.
This year, victims reported losing a median of $10,000 to Bitcoin ATM scams. The FBI warned about this scheme in 2021, and since then, Vermont and Minnesota have enacted laws putting daily transaction limits on crypto kiosks. As is the case with other scams, it’s always important to verify that the email or phone call you receive from a purported company or financial institution is legitimate, and you should never withdraw money based on instructions from a random interaction.
There are other crypto scams on the rise, too, with deepfake crypto scams popping up on YouTube, bad actors impersonating journalists to drain victims’ digital wallets, and “pig butchering” scams that have led to the arrest of a former bank CEO who stole millions to invest in a phony crypto scheme.

Image: Nick Barclay / The Verge

People are losing a lot more money to Bitcoin ATM scams. In a report released on Tuesday, the Federal Trade Commission said it found the amount of money lost to Bitcoin ATM scams increased nearly 10 times from 2020 to 2023 — going from $12 million to a whopping $114 million. Consumers have already lost $65 million to the ruse in the first half of 2024 alone.

Over the years, scams have evolved into different versions of pretty much the same thing: schemes that trick victims into paying scammers. We’ve seen bad actors trick people into sending wire transfers, buying gift cards, and even handing over a pile of cash in a shoebox. This is the Bitcoin ATM variation.

Bitcoin ATMs are the physical machines that you can use to purchase or sell cryptocurrency at places like gas stations and grocery stores. To carry out their scheme, scammers contact a victim by phone, text, or through online pop-ups. They’ll then impersonate a bank or government official while attempting to convince their victim that their bank account has been compromised and that they need to take action.

Image: FTC

Scammers will then instruct their victims to take out a large sum of cash and deposit it into a Bitcoin ATM — which the FTC says scammers may call “safety lockers” — to keep their funds secure. Once at the Bitcoin ATM, scammers will send their victim a QR code to scan at the machine, which will put all of the victim’s cash into the scammer’s crypto wallet.

This year, victims reported losing a median of $10,000 to Bitcoin ATM scams. The FBI warned about this scheme in 2021, and since then, Vermont and Minnesota have enacted laws putting daily transaction limits on crypto kiosks. As is the case with other scams, it’s always important to verify that the email or phone call you receive from a purported company or financial institution is legitimate, and you should never withdraw money based on instructions from a random interaction.

There are other crypto scams on the rise, too, with deepfake crypto scams popping up on YouTube, bad actors impersonating journalists to drain victims’ digital wallets, and “pig butchering” scams that have led to the arrest of a former bank CEO who stole millions to invest in a phony crypto scheme.

Read More 

The Google Pixel 9. | Photo: Allison Johnson / The Verge

Android 15 has officially arrived — for developers, at least. Google’s next-gen operating system is available in the Android Open Source Project, with support for Pixel phones due to roll out in the coming weeks. You’ll have to wait a little longer if you have a compatible device from Samsung, Motorola, OnePlus, Nothing, or other brands, while Google describes the window for Android 15 coming to those phones as within the “coming months.”
While Apple typically rolls out major changes tied directly to each new version of iOS, Google’s Android updates often focus on tweaking the underlying ways apps and devices interact with each other. Down the line, devices with Android 15 will support single-tap passkeys, theft detection, improved multitasking for large-screen devices, and the ability to limit access to certain apps.

Some of the features launching with the system include an update to Android’s screen reader, TalkBack, which Google highlighted at I/O earlier this year. The utility will now have an integration with Google’s Gemini AI assistant to provide audio descriptions of images you’re looking at on your phone.

GIF: Google
Circle to Search is getting a Shazam-like feature that helps you identify songs.

Google is also expanding its Circle to Search feature with a new Shazam-like ability to help you identify songs. After testing the feature over the past several weeks, Google will now let you select a music button to get the name, artist, and YouTube video associated with a song — whether it’s playing from a video you’re watching or grocery store speakers.
Additionally, Google is expanding earthquake alerts to all of the US and six of its territories. The company already partners with the United States Geological Survey (USGS) and its ShakeAlert system to offer warnings in California, Oregon, and Washington using readings from traditional seismometers.
But now, users in states without access to the ShakeAlert system can receive earthquake notifications based on data from their Android device’s accelerometer. If many devices in an area detect vibration, Google says its Android Earthquake Alerts analyze “the crowdsourced data to determine if an earthquake is occurring.” Google has already rolled out this feature in nearly 100 countries.

Image: Google
Google will send these two warnings for earthquakes with a 4.5 magnitude and greater.

Some other features coming to Android devices include a way for Chrome to read webpages aloud as well as support for offline Google Maps in Wear OS 5.

The Google Pixel 9. | Photo: Allison Johnson / The Verge

Android 15 has officially arrived — for developers, at least. Google’s next-gen operating system is available in the Android Open Source Project, with support for Pixel phones due to roll out in the coming weeks. You’ll have to wait a little longer if you have a compatible device from Samsung, Motorola, OnePlus, Nothing, or other brands, while Google describes the window for Android 15 coming to those phones as within the “coming months.”

While Apple typically rolls out major changes tied directly to each new version of iOS, Google’s Android updates often focus on tweaking the underlying ways apps and devices interact with each other. Down the line, devices with Android 15 will support single-tap passkeys, theft detection, improved multitasking for large-screen devices, and the ability to limit access to certain apps.

Some of the features launching with the system include an update to Android’s screen reader, TalkBack, which Google highlighted at I/O earlier this year. The utility will now have an integration with Google’s Gemini AI assistant to provide audio descriptions of images you’re looking at on your phone.

Google is also expanding its Circle to Search feature with a new Shazam-like ability to help you identify songs. After testing the feature over the past several weeks, Google will now let you select a music button to get the name, artist, and YouTube video associated with a song — whether it’s playing from a video you’re watching or grocery store speakers.

Additionally, Google is expanding earthquake alerts to all of the US and six of its territories. The company already partners with the United States Geological Survey (USGS) and its ShakeAlert system to offer warnings in California, Oregon, and Washington using readings from traditional seismometers.

But now, users in states without access to the ShakeAlert system can receive earthquake notifications based on data from their Android device’s accelerometer. If many devices in an area detect vibration, Google says its Android Earthquake Alerts analyze “the crowdsourced data to determine if an earthquake is occurring.” Google has already rolled out this feature in nearly 100 countries.

Image: Google
Google will send these two warnings for earthquakes with a 4.5 magnitude and greater.

Some other features coming to Android devices include a way for Chrome to read webpages aloud as well as support for offline Google Maps in Wear OS 5.

Read More 

The first Pixel Fold (left) and the Pixel 9 Pro Fold (right).

The highest compliment I can pay the Pixel 9 Pro F — oh forget it, let’s just call it the Pixel Fold — is that I forgot I was using a folding phone while I was using it.
Sounds like a weird compliment, right? Isn’t the whole point of the thing that it opens up so you can use the big inner screen? Well, kind of. The whole promise of the foldable phone is that it’s basically a small tablet when you need more screen real estate and a regular phone when you fold it back up. Every foldable I’ve used delivered on the first part of that promise. Big screen? We did it, Joe. It’s the small-screen experience that hasn’t been great.
I was using the phone’s outer screen before bed, drowsily answering the minimum amount of Duolingo questions required to keep my streak intact, and it dawned on me: it just felt like a regular-ass phone.

The Pixel 9 Pro Fold’s longer cover screen (on the right) is a more familiar format than the shorter, wider outer screen on the Pixel Fold.

That’s not an experience I’ve had using Samsung’s Z Fold or the first-generation Pixel Fold. They remind you every second you’re using them that something is up. The Galaxy Z Fold 6 is still too long and skinny to feel like a normal phone, and the first Pixel Fold is so heavy, it feels like holding two phones at once.

This foldable is built different. It’s built like the OnePlus Open, in fact, which has been my favorite book-style foldable ever since I got my hands on it. The previous Pixel Fold was more passport-shaped, with a cover screen that was wider and shorter than a typical slab-style phone. The new Pixel 9 Pro Fold has a 6.3-inch outer screen that’s shaped like the one on the Pixel 9 Pro. Crucially, it’s also much thinner and lighter than the first Fold, so it doesn’t feel like holding a damn brick.
Even just holding the new Fold for the first time felt right. It doesn’t feel like a minor iteration of the first Pixel Fold; it feels like an entirely new phone. That impression hasn’t changed as I’ve used it over the past few days, though I wouldn’t go so far as to say it’s just like carrying a slab phone. It’s still a little bulky in the side pocket of my yoga pants or stuffed into a small outer pocket of my backpack. It’s just not absurdly bulky.

That said, the Pixel Fold still requires you to make some compromises. It’s still $1,799, which is expensive as hell. The cameras still aren’t as good as those of the Pixel 9 Pro or Pro XL. Long-term durability is still a question mark, and one does not simply get a folding phone repaired as easily as a slab-style phone. Foldables have come a long way, but they’re still not quite ready for the mainstream.
I have a lot more testing to do with the, sigh, Pixel 9 Pro Fold. But I’m certain of one thing already: hardware is hard, and this hardware is good.
Photography by Allison Johnson / The Verge

The first Pixel Fold (left) and the Pixel 9 Pro Fold (right).

The highest compliment I can pay the Pixel 9 Pro F — oh forget it, let’s just call it the Pixel Fold — is that I forgot I was using a folding phone while I was using it.

Sounds like a weird compliment, right? Isn’t the whole point of the thing that it opens up so you can use the big inner screen? Well, kind of. The whole promise of the foldable phone is that it’s basically a small tablet when you need more screen real estate and a regular phone when you fold it back up. Every foldable I’ve used delivered on the first part of that promise. Big screen? We did it, Joe. It’s the small-screen experience that hasn’t been great.

I was using the phone’s outer screen before bed, drowsily answering the minimum amount of Duolingo questions required to keep my streak intact, and it dawned on me: it just felt like a regular-ass phone.

The Pixel 9 Pro Fold’s longer cover screen (on the right) is a more familiar format than the shorter, wider outer screen on the Pixel Fold.

That’s not an experience I’ve had using Samsung’s Z Fold or the first-generation Pixel Fold. They remind you every second you’re using them that something is up. The Galaxy Z Fold 6 is still too long and skinny to feel like a normal phone, and the first Pixel Fold is so heavy, it feels like holding two phones at once.

This foldable is built different. It’s built like the OnePlus Open, in fact, which has been my favorite book-style foldable ever since I got my hands on it. The previous Pixel Fold was more passport-shaped, with a cover screen that was wider and shorter than a typical slab-style phone. The new Pixel 9 Pro Fold has a 6.3-inch outer screen that’s shaped like the one on the Pixel 9 Pro. Crucially, it’s also much thinner and lighter than the first Fold, so it doesn’t feel like holding a damn brick.

Even just holding the new Fold for the first time felt right. It doesn’t feel like a minor iteration of the first Pixel Fold; it feels like an entirely new phone. That impression hasn’t changed as I’ve used it over the past few days, though I wouldn’t go so far as to say it’s just like carrying a slab phone. It’s still a little bulky in the side pocket of my yoga pants or stuffed into a small outer pocket of my backpack. It’s just not absurdly bulky.

That said, the Pixel Fold still requires you to make some compromises. It’s still $1,799, which is expensive as hell. The cameras still aren’t as good as those of the Pixel 9 Pro or Pro XL. Long-term durability is still a question mark, and one does not simply get a folding phone repaired as easily as a slab-style phone. Foldables have come a long way, but they’re still not quite ready for the mainstream.

I have a lot more testing to do with the, sigh, Pixel 9 Pro Fold. But I’m certain of one thing already: hardware is hard, and this hardware is good.

Photography by Allison Johnson / The Verge

Read More 

Illustration: Alex Castro / The Verge

Clearview AI has been hit with its largest fine yet under Europe’s General Data Protection Regulation (GDPR) by a Dutch regulator. The Dutch Data Protection Authority (DPA) announced a fine of 30.5 million Euros, or about $33.7 million.
The American facial recognition company, which built a database of images scraped from social media platforms, has been the target of regulators around the world for alleged privacy violations. It’s previously faced fines from the UK, Australia, France, and Italy, and been forced to delete data on those countries’ residents.
The DPA accused Clearview of creating an illegal database with “unique biometric codes” linked to the photos it collected. It also allegedly failed to give people whose faces were in the database sufficient information about how their image and biometric data are used. The company also allegedly continued to violate the law after the Dutch authorities began investigating it, which the DPA said could lead to an additional fine of up to 5.1 million Euros.

Given that Clearview has not changed its behavior after other fines, according to the DPA, the regulator said in a press release it “is looking for ways to make sure that Clearview stops the violations.” Dutch DPA chairman Aleid Wolfsen said in a statement that company directors could be held personally liable if they knew of the GDPR violations and could have stopped them, but chose not to.
The Dutch DPA said Clearview can’t appeal the fine because it hasn’t objected to the decision. But Clearview says it’s not enforceable. In a statement, Clearview’s chief legal officer Jack Mulcaire said in a statement that company, “does not have a place of business in the Netherlands or the EU, it does not have any customers in the Netherlands or the EU, and does not undertake any activities that would otherwise mean it is subject to the GDPR.” Mulcaire said that the decision is “unlawful, devoid of due process and is unenforceable.”

Illustration: Alex Castro / The Verge

Clearview AI has been hit with its largest fine yet under Europe’s General Data Protection Regulation (GDPR) by a Dutch regulator. The Dutch Data Protection Authority (DPA) announced a fine of 30.5 million Euros, or about $33.7 million.

The American facial recognition company, which built a database of images scraped from social media platforms, has been the target of regulators around the world for alleged privacy violations. It’s previously faced fines from the UK, Australia, France, and Italy, and been forced to delete data on those countries’ residents.

The DPA accused Clearview of creating an illegal database with “unique biometric codes” linked to the photos it collected. It also allegedly failed to give people whose faces were in the database sufficient information about how their image and biometric data are used. The company also allegedly continued to violate the law after the Dutch authorities began investigating it, which the DPA said could lead to an additional fine of up to 5.1 million Euros.

Given that Clearview has not changed its behavior after other fines, according to the DPA, the regulator said in a press release it “is looking for ways to make sure that Clearview stops the violations.” Dutch DPA chairman Aleid Wolfsen said in a statement that company directors could be held personally liable if they knew of the GDPR violations and could have stopped them, but chose not to.

The Dutch DPA said Clearview can’t appeal the fine because it hasn’t objected to the decision. But Clearview says it’s not enforceable. In a statement, Clearview’s chief legal officer Jack Mulcaire said in a statement that company, “does not have a place of business in the Netherlands or the EU, it does not have any customers in the Netherlands or the EU, and does not undertake any activities that would otherwise mean it is subject to the GDPR.” Mulcaire said that the decision is “unlawful, devoid of due process and is unenforceable.”

Read More 

Neon

Director Asif Kapadia is best known for his documentaries chronicling the lives of entertainment icons whose work left indelible marks on the worlds of music and sports. But his next feature looks like a nightmarish vision of a future that feels all too possible.
Though 2073, Kapadia’s new film, features plenty of archival news footage from our present, it’s also a sci-fi thriller set in a distant future where years of climate-related disasters and the rise of authoritarianism have transformed the world into a dystopian hellscape. To people like Ghost (Samantha Morton) — a woman barely surviving in an apocalyptic San Francisco by keeping off the grid — it’s all too clear how societies across the globe descended into chaos. But rather than fully dramatize those moments in history, 2073 buttresses its story with real-world newsreels and interviews with journalists about how the gradual erosion of democracy in the past paves the way for a terrifying new world order.
Along with shots of various political strongmen like Vladimir Putin, Narendra Modi, Kim Jong Un, and Donald Trump, 2073’s trailer also emphasizes how Silicon Valley titans like Mark Zuckerberg play key roles in fostering the spread of fascism. And while the film doesn’t yet have a release date, as time-specific as 2073’s story seems to be, it probably won’t be long until it’s playing in theaters.

Neon

Director Asif Kapadia is best known for his documentaries chronicling the lives of entertainment icons whose work left indelible marks on the worlds of music and sports. But his next feature looks like a nightmarish vision of a future that feels all too possible.

Though 2073, Kapadia’s new film, features plenty of archival news footage from our present, it’s also a sci-fi thriller set in a distant future where years of climate-related disasters and the rise of authoritarianism have transformed the world into a dystopian hellscape. To people like Ghost (Samantha Morton) — a woman barely surviving in an apocalyptic San Francisco by keeping off the grid — it’s all too clear how societies across the globe descended into chaos. But rather than fully dramatize those moments in history, 2073 buttresses its story with real-world newsreels and interviews with journalists about how the gradual erosion of democracy in the past paves the way for a terrifying new world order.

Along with shots of various political strongmen like Vladimir Putin, Narendra Modi, Kim Jong Un, and Donald Trump, 2073’s trailer also emphasizes how Silicon Valley titans like Mark Zuckerberg play key roles in fostering the spread of fascism. And while the film doesn’t yet have a release date, as time-specific as 2073’s story seems to be, it probably won’t be long until it’s playing in theaters.

Read More 

Illustration by Cath Virginia / The Verge | Photos from Getty Images

California lawmakers push to regulate artificial intelligence. California is known for taking on regulatory issues like data privacy and social media content moderation, and its latest target is AI. The state’s legislature recently passed SB 1047, one of the US’s first and most significant frameworks for governing artificial intelligence systems. The bill contains sweeping AI safety requirements aimed at the potentially existential risks of “foundation” AI models trained on vast swaths of human-made and synthetic data.
SB 1047 has proven controversial, drawing criticism from the likes of Mozilla (which expressed concern it would harm the open-source community); OpenAI (which warned it could hamper the AI industry’s growth); and Rep. Nancy Pelosi (D-CA), who called it “well-intentioned but ill informed.” But particularly after an amendment that softened some provisions, it garnered support from other parties. Anthropic concluded that the bill’s “benefits likely outweigh its costs,” while former Google AI lead Geoffrey Hinton called it “a sensible approach” for balancing risks and advancement of the technology.
Governor Gavin Newsom hasn’t indicated whether he will sign SB 1047, so the bill’s future is hazy. But the biggest foundation model companies are based in California, and its passage would affect them all.

Illustration by Cath Virginia / The Verge | Photos from Getty Images

California lawmakers push to regulate artificial intelligence.

California is known for taking on regulatory issues like data privacy and social media content moderation, and its latest target is AI. The state’s legislature recently passed SB 1047, one of the US’s first and most significant frameworks for governing artificial intelligence systems. The bill contains sweeping AI safety requirements aimed at the potentially existential risks of “foundation” AI models trained on vast swaths of human-made and synthetic data.

SB 1047 has proven controversial, drawing criticism from the likes of Mozilla (which expressed concern it would harm the open-source community); OpenAI (which warned it could hamper the AI industry’s growth); and Rep. Nancy Pelosi (D-CA), who called it “well-intentioned but ill informed.” But particularly after an amendment that softened some provisions, it garnered support from other parties. Anthropic concluded that the bill’s “benefits likely outweigh its costs,” while former Google AI lead Geoffrey Hinton called it “a sensible approach” for balancing risks and advancement of the technology.

Governor Gavin Newsom hasn’t indicated whether he will sign SB 1047, so the bill’s future is hazy. But the biggest foundation model companies are based in California, and its passage would affect them all.

Read More