Politico reports: Governments and businesses have spent two decades rushing to the cloud — trusting some of their most sensitive data to tech giants that promised near-limitless storage, powerful software and the knowhow to keep it safe.

Now the White House worries that the cloud is becoming a huge security vulnerability.

So it's embarking on the nation's first comprehensive plan to regulate the security practices of cloud providers like Amazon, Microsoft, Google and Oracle, whose servers provide data storage and computing power for customers ranging from mom-and-pop businesses to the Pentagon and CIA.... Among other steps, the Biden administration recently said it will require cloud providers to verify the identity of their users to prevent foreign hackers from renting space on U.S. cloud servers (implementing an idea first introduced in a Trump administration executive order). And last week the administration warned in its national cybersecurity strategy that more cloud regulations are coming — saying it plans to identify and close regulatory gaps over the industry....

So far, cloud providers have haven't done enough to prevent criminal and nation-state hackers from abusing their services to stage attacks within the U.S., officials argued, pointing in particular to the 2020 SolarWinds espionage campaign, in which Russian spooks avoided detection in part by renting servers from Amazon and GoDaddy. For months, they used those to slip unnoticed into at least nine federal agencies and 100 companies. That risk is only growing, said Rob Knake, the deputy national cyber director for strategy and budget. Foreign hackers have become more adept at "spinning up and rapidly spinning down" new servers, he said — in effect, moving so quickly from one rented service to the next that new leads dry up for U.S. law enforcement faster than it can trace them down.

On top of that, U.S. officials express significant frustration that cloud providers often up-charge customers to add security protections — both taking advantage of the need for such measures and leaving a security hole when companies decide not to spend the extra money. That practice complicated the federal investigations into the SolarWinds attack, because the agencies that fell victim to the Russian hacking campaign had not paid extra for Microsoft's enhanced data-logging features.... Part of what makes that difficult is that neither the government nor companies using cloud providers fully know what security protections cloud providers have in place. In a study last month on the U.S. financial sector's use of cloud services, the Treasury Department found that cloud companies provided "insufficient transparency to support due diligence and monitoring" and U.S. banks could not "fully understand the risks associated with cloud services."

An anonymous reader quotes a report from Ars Technica: When you first start messing with the command line, it can feel like there's an impermeable wall between the local space you're messing around in and the greater Internet. On your side, you've got your commands and files, and beyond the wall, there are servers, images, APIs, webpages, and more bits of useful, ever-changing data. One of the most popular ways through that wall has been cURL, or "client URL," which turns 25 this month.

The cURL tool started as a way for programmer Daniel Stenberg to let Internet Chat Relay users quickly fetch currency exchange rates while still inside their chat window. As detailed in an archived history of the project, it was originally built off an existing command-line tool, httpget, built by Rafael Sagula. A 1.0 version was released in 1997, then changed names to urlget by 2.0, as it had added in GOPHER, FTP, and other protocols. By 1998, the tool could upload as well as download, and so version 4.0 was named cURL. Over the next few years, cURL grew to encompass nearly every Internet protocol, work with certificates and encryption, offer bindings for more than 50 languages, and be included in most Linux distributions and other systems. The cURL project now encompasses both the command-line command itself and the libcurl library. In 2020, the project's history estimated the command and library had been installed in more than 10 billion instances worldwide.

How do you celebrate a piece of indispensable Internet architecture turning 25? Stenberg plans to host a "Zoom birthday party" at 17:00 UTC time on March 20. Double-check that time in your area: "It is within this weird period between [when] the US has switched to daylight saving time while Europe has not yet switched," Stenberg writes on his blog. Stenberg plans to sip on a 25-year Bowmore Islay single-malt Scotch, while presenting the project's history and future plans while taking questions. (A link to the Zoom call will be added to Stenberg's blog post closer to March 20.)

Cerebral has revealed it shared the private health information, including mental health assessments, of more than 3.1 million patients in the United States with advertisers and social media giants like Facebook, Google, and TikTok. From a report: The telehealth startup, which exploded in popularity during the COVID-19 pandemic after rolling lockdowns and a surge in online-only virtual health services, disclosed the security lapse in a filing with the federal government that it shared patients' personal and health information who used the app to search for therapy or other mental health care services. Cerebral said that it collected and shared names, phone numbers, email addresses, dates of birth, IP addresses and other demographics, as well as data collected from Cerebral's online mental health self-assessment, which may have also included the services that the patient selected, assessment responses, and other associated health information.

Belgian federal government employees will no longer be allowed to use the Chinese-owned video app TikTok on their work phones, Belgian Prime Minister Alexander De Croo said on Friday. From a report: De Croo said the Belgian national security council had warned of the risks associated with the large amounts of data collected by TikTok, which is owned by Chinese firm ByteDance, and the fact that the company is required to cooperate with Chinese intelligence services. "That is the reality," the prime minister said in a statement. "That's why it is logical to forbid the use of TikTok on phones provided by the federal government. The safety of our information must prevail." The European Commission and the European Parliament last month banned TikTok from staff phones due to growing concerns about the company, and whether China's government could harvest users' data or advance its interests.

Twitter's decline is paving the way for other platforms to build next-generation replacements. And now the biggest player in the game is getting involved: Meta is in the early stages of building a dedicated app for people to post text-based updates. From a report: "We're exploring a standalone decentralized social network for sharing text updates," the company told Platformer exclusively in an email. "We believe there's an opportunity for a separate space where creators and public figures can share timely updates about their interests." News that Meta has been exploring a text-based network was first reported Thursday by MoneyControl. The app is codenamed P92 and will allow users to log in through their existing Instagram credentials, the outlet reported.

Details about the project are scant. The product is still in its earliest stages, sources said, and there is no time frame for it being released. But legal and regulatory teams have already started to investigate potential privacy concerns around the app so they can be addressed before launch, we're told. Adam Mosseri, who runs Instagram, is taking the lead on the project, sources said. The most remarkable aspect of the project is that Meta plans for the network to be decentralized. While the company would not elaborate beyond its statement, in a decentralized network individual users are typically able to set up their own, independent servers and set server-specific rules for how content is moderated. Building a decentralized network could also give Meta the opportunity for its new app to interoperate with other social products -- a previously unheard-of gesture from a company known for building some of the most lucrative walled gardens in the industry's history.

According to Bloomberg, Google wants to build AI into everything to fight OpenAI's ChatGPT. Google issued "a directive that all of its most important products -- those with more than a billion users -- must incorporate generative AI within months."

Ars Technica's Ron Amadeo likens it to the company's failed Google+ playbook from 2011. To combat Facebook's rising popularity, then-Google CEO Larry Page directed employees to build social features into everything. YouTube comments were tied to Google+, Gmail addresses required a Google+ account, Google Search had "+1" buttons, and a "real name" policy was instituted, among other things. "That forced integration strategy was an abject failure, and after a few years of Google's social panic, all of Google+'s integrations were removed, and the service was eventually shut down," writes Amadeo. An anonymous reader shares an excerpt from Amadeo's report: We wrote last month that Google's ChatGPT panic seemed a lot like its response to Google+, and several employees relayed that same sentiment to Bloomberg. Just like with G+, the report added that "current and former employees say at least some Googlers' ratings and reviews will likely be influenced by their ability to integrate generative AI into their work."

AI is one of the few areas of Google that CEO Sundar Pichai is really invested in, with the CEO saying the technology would be "more profound than fire or electricity." Google was, for years, a leader in AI with voice recognition features like the Google Assistant, speech synthesis features like Google Duplex, and mastering the game of Go. Those features debuted years ago, though, and a fear of rolling out imperfect products has meant Google locks a lot of technology away in a lab somewhere. In a 2021 New York Times article that was critical of Pichai's management style, "A common critique among current and former executives is that Mr. Pichai's slow deliberations often feel like a way to play it safe and arrive at a 'no.'" Despite many seeing Pichai as the source of Google's reluctance, the Bloomberg report says the CEO is now taking a more hands-on approach to product development, saying, "The effort has Pichai reliving his days as a product manager, as he's taken to weighing in directly on the details of product features, a task that would usually fall far below his pay grade, according to one former employee."

As for exactly what these forced AI integrations will look like, the report cites a recent YouTube feature that would let people virtually swap outfits. In Alphabet's Q4 2022 (PDF) earnings call, Pichai said the company was "working to bring large language models to Gmail and Docs," so expect to be able to click a few buttons soon and have those apps generate blocks of text. The Bloomberg article quotes one Google employee as saying, "We're throwing spaghetti at the wall, but it's not even close to what's needed to transform the company and be competitive."

Alphabet's Wing is debuting a Wing Delivery Network platform that relies on decentralized and highly automated pickups. Engadget reports: Drones charge and deliver in whatever locations make the most sense for the broader system. If demand surges in a given area, more drones can operate around the nearest pads. Crucially, your local restaurant or store doesn't have to do much to take advantage of the network. An AutoLoader system lets shops simply latch a package to a curbside pickup location and walk away -- the drone handles the rest. Businesses have to order drones, but they don't have to manage the fleet or make employees wait for an aircraft to arrive.

The technology is also meant to scale elegantly. It's relatively easy to add new pad locations as usage grows, and the drones can double as scouts that expand the network. The drones can even make sure they're allowed to fly in a given area. Wing expects "elements" of the Delivery Network to deploy over the next year, with demonstrations taking place worldwide in 2023. Provided all goes according to plan, the brand wants to handle "millions" of deliveries by mid-2024, at prices that beat conventional ground-based delivery.

The new 2024 Ford Mustang is losing its AM radio receiver, reports The Drive. A Ford spokesperson confirmed the feature's deletion, citing that "countries and automakers globally are modernizing radio by offering internet streaming through mobile apps, FM, or digital." From the report: The availability of AM radio in new cars has declined over the last decade with shifts in media consumption habits, with AM radio's audience evaporating in Europe and accounting for only a small minority of the U.S. population. [...] It's a trend of concern to current and former U.S. emergency officials, who recently wrote the Secretary of Transportation a letter advising action on AM radio's disappearance. AM radio is a crucial component of the U.S.'s national alert network, with just 75 stations reaching more than 90 percent of the country's population. Officials are concerned that AM receivers' decreasing availability in new cars could compromise their ability to reach citizens during emergencies.

An anonymous reader shares a report: Two years ago, I wrote a reasonably prescient editorial about how the writing was on the wall for Google's cloud gaming service Stadia -- and how the company was now hoping to sell its white label streaming technology to other companies instead of building out its own Netflix of games. But it seems that, when Google killed off Stadia, it threw away that technology, too. Google executive Jack Buser has now admitted that the company is no longer offering the white label version of Stadia that allowed companies like AT&T and Capcom to let anyone try games like Batman: Arkham Knight, Control, and a demo of Resident Evil Village for free over the internet, not to mention the first game from Peloton.

"We are not offering that streaming option, because it was tied to Stadia itself," he told Axios' Stephen Totilo. "So unfortunately, when we decided to not move forward with Stadia, that sort of offering could no longer be offered as well." Google called the white label version "Immersive Stream for Games" and sometimes "Google Stream" and, to my knowledge, it was only ever used in experiments like the ones I link above. In AT&T's case, they were limited to its own internet subscribers. Maybe they weren't that successful? When we spoke to AT&T about cloud gaming following those experiments, the carrier didn't seem that bullish about serving up more games itself.

Global shutter sensors with no skew or distortion have been promised as the future of cameras for years now, but so far only a handful of products with that tech have made it to market. Now, Raspberry Pi is offering a 1.6-megapixel global shutter camera module to hobbyists for $50, providing a platform for machine vision, hobbyist shooting and more. From a report: The Raspberry Pi Global Shutter Camera uses a 6.3mm Sony IMX296 sensor, and requires a Raspberry Pi board with a CSI camera connector. Like other global shutter sensors, it works by pairing each pixel with an analog storage element, so that light signals can be captured and stored by all pixels simultaneously. By comparison, regular CMOS sensors read and store the light captured by pixels from top to bottom and left to right. That can cause diagonal skew on fast moving subjects, or very weird distortion on rotating objects like propellers.

schwit1 shares an excerpt from a Substack article, written by former cybersecurity reporter Catalin Cimpanu: The Canada Revenue Agency (CRA), the tax department of Canada, recently updated its terms and conditions to force taxpayers to agree that CRA is not liable if their personal information is stolen while using the My Account online service portal -- which, ironically, all Canadians must use when doing their taxes and/or running their business. The CRA's terms of use assert the agency is not liable because they have "taken all reasonable steps to ensure the security of this Web site."

Excerpt from the CRA terms statement: "10. The Canada Revenue Agency has taken all reasonable steps to ensure the security of this Web site. We have used sophisticated encryption technology and incorporated other procedures to protect your personal information at all times. However, the Internet is a public network and there is the remote possibility of data security violations. In the event of such occurrences, the Canada Revenue Agency is not responsible for any damages you may experience as a result."

Unfortunately, that is not true. After reviewing the HTTP responses from the CRA My Account login page, it's clear the agency has not configured even some of the most basic security features. For example, security protections for their cookies are not configured, nor are all the recommended security headers used. Not only is that not "all reasonable steps," but the CRA is missing the very basics for securing online web applications.

The terms of use also state that users are not allowed to use "any script, robot, spider, Web crawler, screen scraper, automated query program or other automated device or any manual process to monitor or copy the content contained in any online services." Looking at the HTTP response headers using web browser developer tools doesn't breach the terms of services, but the CRA must be well aware that internet users perform scans like this all the time. And it's not the legitimate My Account users who are likely to be the culprits. Unfortunately for Canadians, threat actors don't read terms of use pages. A statement like this doesn't protect anyone, except CRA, from being held responsible for failing to properly secure Canadian citizens' personal data.

Microsoft is starting to test a system called File Recommendations in File Explorer, which does exactly what the name suggests -- when you visit the home tab, it shows specific files that you may want to open at the top. The Verge reports: In a blog post, the company says the current version is only available to some Insiders in its Dev Channel who have installed the Windows 11 Insider Preview Build 23403 update and will only work if you're logged in with an Azure Active Directory account (meaning that currently, this feature feels squarely aimed at business users). For those that do have it, it'll suggest cloud files that you own or that have been shared with you.

Microsoft says it plans to "monitor feedback and see how it lands before pushing it out to everyone," so it seems as if it's aware that the feature could be controversial. Part of that may be just down to the fact that not everybody will want unexpected results in their file browser -- though based on the screenshot, you will be able to collapse the Recommended section.

Last month, Volkswagen garnered plenty of bad publicity when it emerged that the company's connected car service refused to help track a stolen car -- with a 2-year-old child still on board -- until someone paid to reactivate the service. Now, the automaker says it's very sorry this happened, and it's making its connected vehicle emergency service free to most model-year 2020-2023 Volkswagens. Ars Technica reports: "The family was thankfully reunited, but the crime and the process failure are heartbreaking for me," said Rachael Zaluzec, VW's SVP for customer experience and brand and marketing. "As a mom and an aunt, I can imagine how painful this incident must have been. Words can't adequately express how truly sorry I am for what the family endured."

"Volkswagen must and will do better for everyone that trusts our brand and for the law enforcement officials tasked with protecting us. In addition to a full investigation of what went wrong and actions taken to address the failure, we want to make it right for the future. Today, we are setting a new standard for customer peace of mind. As of June 1, we will make these connected vehicle emergency services free for five years as one significant step we can take as a commitment to our owners and their families," Zaluzec said in a statement sent to Ars.

Most MY2020 or newer VWs can use connected services, apart from MY2020 Passats. From June, owners can sign up for five years of free Car-Net Safe and Secure, which uses the vehicle's onboard modem to connect to the emergency services via the car's SOS button. In gasoline-powered VWs, there is also an anti-theft alert. VW says it will make Car-Net Remote Access free for five years as well. This lets owners interact with their car via a mobile app and can lock and unlock the doors, honk the horn and flash the lights, and, if fitted, remote-start the vehicle.

An anonymous reader quotes a report from Politico: The Dutch government confirmed for the first time Wednesday it will impose new export controls on microchips manufacturing equipment, bowing to U.S. pressure to block the sale of some of its prized chips printing machines to China. The U.S. and the Netherlands reached an agreement to introduce new export restrictions on advanced chip technology to China at the end of January, but until now, the Dutch government hadn't commented publicly on it. The deal, which also included Japan, involves the only three countries that are home to manufacturers of advanced machines to print microchips. It is a U.S.-led initiative to choke off the supply of cutting-edge chips to China.

"Given the technological developments and geopolitical context, the government has concluded that it is necessary for the (inter)national security to expand the existing export controls on specific manufacturing equipment for semiconductors," Foreign Trade Minister Liesje Schreinemacher wrote in a letter to Dutch lawmakers published Wednesday evening. The Dutch government wants to prevent Dutch technology from being used in military systems or weapons of mass destruction, Schreinemacher wrote — echoing the U.S. reasoning when it imposed its own export controls in October. The Netherlands also wants to avoid losing its pole position in producing cutting-edge chip manufacturing tools: Schreinemacher said the government wants to uphold "Dutch technological leadership." While China is not explicitly named in Schreinemacher's letter, the new policy is targeted at Chinese efforts to overtake the U.S. and others like Taiwan, South Korea, Japan and leading European countries in the global microchips supply chain.

The new export restrictions deal a blow to ASML, the global leader in producing advanced microchips printing machines based in Veldhoven, in southern Netherlands. In the letter, Schreinemacher said the new export control measures include the most advanced deep ultraviolet (DUV) machines, which are part of ASML's advanced chips printers portfolio. The Dutch firm, which is the highest-valued tech company in Europe, already did not receive export licenses for selling its most advanced machines using extreme ultraviolet light (EUV) technology to China since 2019. ASML in a statement confirmed it will now "need to apply for export licenses for shipment of the most advanced immersion DUV systems," but it noted it has not yet received more details about what "most advanced" means.

An anonymous reader quotes a report from the Associated Press: In the past month, bed and breakfast owner Chen Yu-lin had to tell his guests he couldn't provide them with the internet. Others living on Matsu, one of Taiwan's outlying islands closer to neighboring China, had to struggle with paying electricity bills, making a doctor's appointment or receiving a package. For connecting to the outside world, Matsu's 14,000 residents rely on two submarine internet cables leading to Taiwan's main island. The National Communications Commission, citing the island's telecom service, blamed two Chinese ships for cutting the cables. It said a Chinese fishing vessel is suspected of severing the first cable some 50 kilometers (31 miles) out at sea. Six days later, on Feb. 8, a Chinese cargo ship cut the second, NCC said.

Taiwan's government stopped short of calling it a deliberate act on the part of Beijing, and there was no direct evidence to show the Chinese ships were responsible. The islanders in the meantime were forced to hook up to a limited internet via microwave radio transmission, a more mature technology, as backup. It means one could wait hours to send a text. Calls would drop, and videos were unwatchable. "A lot of tourists would cancel their booking because there's no internet. Nowadays, the internet plays a very large role in people's lives," said Chen, who lives in Beigan, one of Matsu's main residential islands.

Apart from disrupting lives, the loss of the internet cables, seemingly innocuous, has huge implications for national security. As the full-scale invasion of Ukraine has shown, Russia has made taking out internet infrastructure one of the key parts of its strategy. Some experts suspect China may have cut the cables deliberately as part of its harassment of the self-ruled island it considers part of its territory, to be reunited by force if necessary. The cables had been cut a total of 27 times in the past five years, but it was unclear which country the vessels hailed from, based on data from Chunghwa Telecom.

An anonymous reader shares a blog: Google Groups is dying. Its epitaph is not yet inscribed on the Killed by Google website, but the end is easily seen from here (although it should also be noted its death was called as early as 13 years ago). The deficiencies in Google Groups search, supposedly Google's forte, have long been noted. Lately though basic features have just stopped working. Why care? Google Groups is, for whatever reason, the de-facto standard community website in the formal methods community. TLA+ uses it, PRISM uses it, SMT-LIB uses it, and a number of other tools I could find at least have presences on the platform. These communities take time to build: their value resides in the number of people who think of them first whenever they want to ask a question or just talk about these tools. Many websites link to these groups or to specific answers.

What's broken? Beyond search's perpetual brokenness, Monospace fonts have just stopped rendering. This makes code samples significantly more difficult to read. Then (precipitating this post) messages or replies submitted to the group have started intermittently just self-deleting. Not in a graceful way, either: the post will successfully be submitted, but then all that will show up is a deleted message. Hope you didn't spend twenty minutes typing a thorough, detailed response! My attempt to submit a link to this very post also deleted itself.

Google is expanding VPN access to all Google One members on all plans and rolling out a new dark web report feature for all subscribers. From a report: VPN by Google One was previously only available to members on the Premium 2TB plan, but will now be available to all Google One members, including those on the Basic plan that starts at $1.99 per month. The tech giant notes that VPN by Google One adds more protection to your internet activity no matter what apps or browsers you use, shielding it from hackers or network operators by masking your IP address. Google is also introducing a new feature called "dark web report" for Google One members on all plans in the United States to help users monitor their personal information on the dark web. Dark web report will start rolling out over the next few weeks to members across all Google One plans in the United States.

As media executives look to pop the stock of flagging publicly traded companies, tech advances are becoming the new gimmick to wow (even temporarily) the investing class. From a report: When Endeavor CEO Ari Emanuel opened his company's earnings call with prepared remarks Feb. 28, a casual listener hearing him tout the UFC and WME might have missed the most interesting part: Emanuel wasn't speaking at all. Well, he sort of was. The words were his, and the voice was his, but rather than Emanuel speaking into a microphone (opening remarks on earnings calls are often pretaped), the comments were the product of a generative artificial intelligence firm calledâSpeechify. Indeed, artificial intelligence has been hard to avoid for Wall Street watchers this year. And it's easy to understand why: AI news has led to share price surges at companies like BuzzFeed (which uses it and saw its stock price more than double) and Microsoft (whose rise was in the high singleâdigits).

After the public release of OpenAI's large language model (LLM) chatbot ChatGPT (and the stock bounce it gave to Microsoft, which is OpenAI's tech partner), it seems that every company wants a piece of the AI action, or at least wants to send the message that it's thinking about it. Companies with close ties to media and entertainment are no exception. In his first letter to creators as CEO of YouTube, Neal Mohan wrote March 1 that "the power of AI is just beginning to emerge in ways that will reinvent video and make the seemingly impossible possible," and that it will be a priority for him. Spotify, led by CEO Daniel Ek, released on Feb. 22 what it is calling an "AI DJ," powered by technology from OpenAI. The DJ takes the music a Spotify user listens to and combines it with recommendations from music experts and an AI-generated voice to create a totally personalized radio show. Snap, run by Evan Spiegel, released an assistant called "My AI," based on ChatGPT, bringing the capabilities of that LLM to Snapchat+ users.

And at small digital publishers like the Jonah Peretti-led BuzzFeed and Arena Group Holdings (the owner of Sports Illustrated), AI is being touted as a fount of new types of storytelling that "can create enterprise value for our brands and partners," per Arena Group CEO Ross Levinsohn. To be sure, AI's potential for transforming business is real (in a March 2 research report, Morgan Stanley called it a "$6 trillion opportunity"), but it remains just that: an opportunity, rather than today's reality. Wall Street is still very bullish on AI in the long term, with Bank of America's Haim Israel and Martyn Briggs writing in a Feb. 28 thematic report that AI is "at a defining moment -- like the internet in the '90s," but the consensus is that while big tech firms like YouTube owner Alphabet, Amazon and Apple will reap the rewards at some point, what it means for smaller companies in the present is less clear. For entertainment companies, the potential is obvious, even if the business models are not. The Morgan Stanley report noted the logic in AI recommendations on streaming services like Spotify and Netflix (Spotify's DJ is a first step in that direction), and it isn't a stretch to think that content itself can be made faster and at less cost with generative technology (applications could include special effects, which are labor intensive and costly). For companies like Endeavor and CAA, generative voice technology (like that from Speechify) and deepfake tech (like Deep Voodoo, a deepfake company from the creators of South Park that counts the CAA-affiliated Connect Ventures as an investor), could mean new opportunities for old actors (see Robert Zemeckis' upcoming movie Here, which will use AI from Metaphysic to de-age Tom Hanks).

Google is warning employees that fewer of them will receive promotions to more senior levels this year than in the past. From a report: "The process is manager-led and will be largely similar to last year -- though with our slower pace of hiring, we are planning for fewer promotions into L6 and above than when Google was growing quickly," the company said in an email that was viewed by CNBC. The L6 distinction refers to the first layer of staff that's considered senior and typically includes people with about a decade of experience. The changes come as Google implements a new performance review system called Google Reviews and Development (GRAD), which as CNBC reported in December, will result in more Google employees receiving low performance ratings and fewer receiving high marks.

Like many large tech companies, Google has a sprawling middle management. According to last year's internal survey results that affected the company's ability to ship products efficiently. Google also is in the midst of trying to cut costs as growth decelerates and recession concerns persist. The company has slowed hiring and announced in January that it's cutting 12,000 jobs, or about 6% of the workforce. In Monday's email, the tech giant said it's promoting fewer people to senior roles "to ensure that the number of Googlers in more senior and leadership roles grows in proportion to the growth of the company."

Binance, the world's largest crypto exchange by trading volume, has extended its spot market share across crypto exchanges for a fourth consecutive month. From a report: The exchange market share increased from 59.4% in January to 61.8% in February, according to a report from crypto market data provider CryptoCompare. Binance had a 13.7% increase in its spot volumes to $504 billion, an all-time high market share for the exchange. This comes as regulators in the U.S. and beyond have ratcheted up their scrutiny of the exchange in recent months. Most recently, a U.S. Securities and Exchange (SEC) official said that agency staff believe Binance.US may be operating an unregistered securities exchange in the U.S., an assertion to which Binance.US objected. Binance is a corporate entity that operates in US through Binance.US.

"Despite the recent criticism the exchange has received, market participants continue to take shelter on Binance under the premise that the largest exchange is seen as one of the safer trading venues," said Jacob Joseph, a research analyst at CryptoCompare, in an interview with CoinDesk. Joseph also attributes the exchange's dominance to the vast amount of liquidity available on Binance, which means reduced slippage costs and spreads, an attractive benefit for traders. "It is one of the exchanges with the most trading pairs and services available," Joseph said.