Connect with us

Tech

AI-Powered Cybercrime Is Surging. The US Lost $16.6 Billion in 2024.

Published

on

I was fortunate enough to spend several days last week at the Aspen Institute’s Crosscurrent summit on AI and national security in San Francisco. My first takeaway: I very much recommend being in sunny (at the moment, at least) San Francisco rather than slushy, raw New York in early March. The second took a little longer to form.

The conference was full of former national security officials, cybersecurity executives, and AI leaders, and the conversation mostly went where you’d expect: the Anthropic-Pentagon fight, the role of AI in the Iran conflict, the coming of autonomous weapons. But the panel that stuck with me was about something less dramatic. It was about something almost old-fashioned, now supercharged by AI: scams.

At one point, Todd Hemmen, a deputy assistant director in the FBI’s Cyber Division’s Cyber Capabilities branch, described how North Korean operatives are using AI-generated face overlays to pass remote job interviews at Western tech companies — then working multiple remote positions simultaneously, funneling the salaries and any intelligence back to the regime in Pyongyang. They fabricate résumés with AI, prep for interviews with AI, and use AI to wear the “face of someone who’s not the person behind the camera,” Hemmen told the audience. Some of the most proficient actors are holding down several full-time jobs at once, all under fake identities, all enabled by tools that didn’t exist two years ago.

That detail has been rattling around in my head since, not the least because it made me wonder how these industrious operatives can manage multiple jobs when I find just one taxing enough. But Hemmen’s story captures something deeper about the moment we find ourselves in. The AI risks getting the most airtime right now are speculative and cinematic — killer robots, AI panopticons. But the AI threat that’s here right now is a foreign agent wearing a synthetic face on a Zoom call, collecting a paycheck from your company. And almost nobody is treating it with the same urgency.

Advertisement

How cybercrime got worse than ever

Cybercrime has been a problem since the days of dial-up, but the scale of what’s happening now is staggering. The FBI reported that the US suffered $16.6 billion in known cybercrime losses in 2024 — up 33 percent in a single year, and more than doubled over three years. Americans over 60 lost nearly $5 billion. And those are just the reported numbers; Alice Marwick, director of research at Data & Society, told the Aspen Institute audience that only about one in five victims ever reports a scam. The real number is unknowable, but it’s much worse.

And now comes generative AI to make all of this faster, cheaper, and more convincing. Phishing emails no longer arrive riddled with typos from supposed Nigerian princes; LLMs can produce fluent, regionally specific language. AI image generators can create entire synthetic identities — dozens of photos of a person who doesn’t exist, complete with vacation shots and designer handbags.

Voice cloning has enabled heists that were science fiction five years ago: In early 2024, a finance worker at the Hong Kong office of UK engineering firm Arup transferred $25 million after a deepfake video call in which the company’s CFO and several colleagues seemed to appear on screen. All of them, it turns out, were fake. CrowdStrike’s 2026 Global Threat Report found that AI-enabled attacks surged 89 percent year-over-year, while the average time from initial breach to being able to spread throughout a network dropped to just 29 minutes. The fastest observed breakout: 27 seconds.

Advertisement

Will AI cyberoffense beat AI cyberdefense?

Why is this problem so comparatively neglected? Partly because we’ve normalized it. Cybercrime has been growing for years, driven by the professionalization of criminal syndicates, cryptocurrency, remote work, and the industrialization of scam compounds in Southeast Asia. (My Vox colleague Josh Keating wrote a great story a couple of years ago on these so-called pig butchering scams.)

We’ve absorbed each year’s record losses as the cost of doing business online. But the curve is steepening: Deloitte projects that generative AI-enabled fraud losses in the US alone could hit $40 billion by 2027. “In the same way that legitimate businesses are integrating automation, so are organized crime,” Marwick said.

That so much of this goes unsaid and unreported adds to the toll. Marwick’s research focuses on romance scams — people targeted during periods of loneliness or transition, slowly bled of their savings by someone they believe loves them. She told the audience that victims often refuse to believe they’re being scammed even when confronted with direct proof. AI makes the emotional manipulation far more persuasive, and no spam filter will protect someone who is willingly sending money.

Advertisement

Can defense keep up? Marwick drew a hopeful comparison to spam, which nearly broke email in the 1990s before a combination of technical fixes, legislation, and social adaptation tamed it, at least to a large extent. Financial institutions are deploying AI to catch AI-enabled fraud. The FBI froze hundreds of millions in stolen funds last year.

But the consensus at the conference was largely grim. “We’re entering this window of time where the offense is so much more capable than the defense,” said Rob Joyce, former director of cybersecurity at the National Security Agency. Marwick was blunter: “I would say generally I’m pretty pessimistic.”

So am I. As I was writing this story, I received an email from a friend with what appeared to be a Paperless Post invitation. The language in the email looked a little odd, but when I clicked on the invite, it took me to a page that seemed very similar to Paperless Post, down to the logo. Still suspicious, I emailed my friend, asking if this was real. “Yes, it is legit,” he wrote back.

That was enough proof for me, but I got distracted and didn’t click on the next step of the invite. Good thing — a few minutes later, my friend emailed me and others to tell us that, yes, he had been hacked.

Advertisement

A version of this story originally appeared in the Future Perfect newsletter. Sign up here!

Source link

Advertisement
Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tech

Everything we know so far

Published

on

The Steam Machine is back from the dead. Not as a Valve-supported program for manufacturers to create living room PCs, but instead as a home console sibling to the Steam Deck. Valve introduced its second attempt at ruling the living room in a surprise hardware announcement in November 2025, and paired the new Steam Machine with a new Steam Controller and a wireless VR headset it calls the Steam Frame. Since the announcement, as is often the case with Valve, some details remain elusive, however.

While we wait for the release of the company’s new hardware lineup in 2026, and more information straight from the horse’s mouth, here’s everything we know about the hardware, software and price of the Steam Machine, so far.

What’s the Steam Machine’s hardware like?

A line-drawing diagram of the Steam Machine and its various ports.

Valve

Like the Steam Deck, the Steam Machine is utilitarian and bespoke. The PC is a black, 5.98 x 6.39 x 6.14 inch (152 x 162.4 x 156mm) box, with ports and a grille for a fan in the back and a removable faceplate and customizable LED light strip in the front. Inside, Valve says the Steam Machine features a “semi-custom” AMD Zen 4 CPU with six cores and up to 4.8GHz clock speeds, and a “semi-custom” RDNA3 AMD GPU, along with 16GB DDR RAM, 8GB GDDR6 VRAM and either 512GB or 2TB of storage.

While these specs make the Steam Machine more powerful than the aging Steam Deck (which shipped in 2022 with its own custom AMD chip) Valve has been careful not to oversell the capabilities of the box. In a blog post, the company said that “the majority of Steam titles play great at 4K 60FPS” using AMD’s FidelityFX Super Resolution (FSR) frame generation and upscaling technology, but some titles require more upscaling than others, and it “may be preferable to play at a lower framerate with [variable refresh rate] to maintain a 1080p internal resolution.”

Advertisement

In a hands-on preview of the Steam Machine, Digital Foundry expressed concern with what Valve’s claims and the device’s stated specs could mean for future performance. “The decision to opt for 8GB of GDDR6 memory has been proven to be a limiting factor on many modern mainstream triple-A games and falls short of the maximum VRAM pools and memory bandwidth available on both Xbox Series X and base PS5,” Digital Foundry writes.

The Steam Machine supports Bluetooth 5.3, Wi-Fi 6E and includes an integrated 2.4GHz adapter for the new Steam Controller. In terms of port selection, there’s DisplayPort 1.4 and HDMI 2.0 inputs for connecting the box to external monitors and TVs, four USB-A ports (divided between two USB 2.0 ports and two USB 3.2 Gen 1 ports) and one USB-C port on the back.

Engadget will have to try out the Steam Machine to really know what it’s capable of, but there’s nothing to suggest it couldn’t be as flexible as the Steam Deck, especially with more power to play with.

What games will be able to run on the Steam Machine?

A Steam Machine connected to a TV playing the game Cuphead.

Valve

Any game that runs on SteamOS, Valve’s Linux-based operating system, will run on the Steam Machine, provided the device’s technical specs will support it. For games running natively on Linux, the Steam Machine will download the Linux version. For Windows games and everything else, it’ll be able to use Steam’s built-in Proton compatibility layer to translate games to Linux, just like the Steam Deck does.

Advertisement

Proton is developed by both Valve and CodeWeavers, the team behind the macOS compatibility app CrossOver. Valve’s compatibility layer translates a game’s API calls and other software features into something Linux understands, essentially tricking the game into thinking it’s running on Windows when it isn’t. Proton has worked remarkably well so far, in some cases helping some PC games run more efficiently on Linux than they do on Windows, but it does have some limitations. Because some anti-cheat software doesn’t support Linux, many competitive multiplayer games aren’t playable on SteamOS. Valve hopes the Steam Machine will help change that.

“While [the] Steam Machine also requires dev participation to enable anti-cheat, we think the incentives for enabling anti-cheat on Machine to be higher than on Deck as we expect more people to play multiplayer games on it,” Valve told Eurogamer. “Ultimately we hope that the launch of Machine will change the equation around anti-cheat support and increase its support.”

To help users find what games work well on the Steam Machine, Valve plans to expand its program for verifying games on the Steam Deck to include the Steam Machine and Steam Frame. Valve looks at things like controller support, the default resolution of the game, whether or not it requires a separate launcher and whether the game and its middleware work with Proton to determine a game’s rating. Then the company sorts games into four categories: Verified (where the game works with Steam hardware at launch), Playable (where a user might have to make modifications to run smoothly), Unplayable (where some or all of the game doesn’t function) and Unknown.

A slide from Valve's GDC 2026 presentation, going over its expectations for Steam Machine games.

Valve

According to an announcement Valve sent to developers in November 2025, games that were Verified for the Steam Deck will automatically be verified for the Steam Machine. In a presentation at GDC 2026, the company also shared that Steam Machine Verified games will be expected to support the same input methods as the Steam Deck and run at 1080p at 30fps at a minimum. Unlike the company’s handheld, Valve won’t require developers to support specific display resolutions or meet legibility requirements to be Steam Machine Verified, though, because the Steam Machine is more likely to be connected to larger displays. That means a game could be marked as Playable on the Steam Deck due to its small text, but Verified on the Steam Machine.

Advertisement

Valve’s system is helpful, but far from definitive — some Unplayable games are in fact playable on the Steam Deck — which is why online, community-run databases like ProtonDB fill in the gaps with more granular information.

How much will the Steam Machine cost and when will it launch?

A Steam Machine with an LED strip displaying the current download progress of a game.

Valve

Valve hasn’t announced a price or a release date for the Steam Machine or any of its new hardware, beyond affirming its new hardware will ship in 2026. In terms of price, however, the company has suggested it might not be a deal in quite the same way the $399 Steam Deck LCD was. Valve designer Pierre-Loup Griffais told The Verge that the “Steam Machine’s pricing is comparable to a PC with similar specs” and that its price would be “positioned closer to the entry level of the PC space” but be “very competitive with what you a PC you could build yourself from parts.”

That means the Steam Machine will likely cost more than the $499 PS5, and that the rising costs of memory could make it even more expensive. Valve has already publicly admitted that memory and storage shortages are affecting its plans. In February, the company said that it was delaying the launch of its hardware (though it still hopes to ship in the first half of 2026) and rethinking pricing, particularly around the Steam Machine and Steam Frame, because of the “limited availability and growing prices” of critical components like RAM.

The changes Framework had to make to the pricing of the Framework Desktop are an illustrative example of the position Valve is in. Framework pitched its compact desktop PC as being great for gaming, with an AMD Ryzen AI Max chip (originally meant for gaming laptops) and a minimum of 32GB of RAM that lets it run games at 1440p. The company originally sold the base configuration of the Framework Desktop for $1,099, but announced in January 2026 that it would now cost $1,139 due to the rising cost of RAM. The price situation got even worse for configurations with more RAM. A Framework Desktop with 128GB of RAM now costs $2,459.

Advertisement

The blame for rising costs lies squarely with the AI industry, whose demand for RAM has led to the collapse of consumer RAM brands and a dearth of true deals on the in-demand component. At this point, PC makers have no solution to the problem other than riding the shortage out and raising prices. Valve clearly isn’t immune to those same issues.

That doesn’t rule out the company offering its Linux PC at multiple different price points, or in some kind of bundle deal with multiple pieces of new Steam hardware. But it does mean that the Steam Machine will likely be priced like a premium device. Same for the Steam Controller and Steam Frame. In the case of the Frame, UploadVR reports that Valve wants to sell the headset for less than the $1,000 Valve Index, but that doesn’t mean it won’t be significantly more expensive than the $300 Meta Quest 3S.

What accessories will work with the Steam Machine?

Someone holding a Steam Controller in a pile of plushies.

Valve

The Steam Machine is designed to work with a variety of different Bluetooth controllers and other wireless accessories, and also whatever you can plug into its multiple USB-A ports and single USB-C port. With a built-in 2.4GHz Steam Controller dongle inside the Steam Machine, Valve’s controller should be an ideal option for controlling games, particularly because of its multiple input options, like touchpads and gyroscopes. Support for Steam Link, Valve’s tech for streaming PC games over local wireless, means you can also send games from a Steam Machine to the Steam Deck, Steam Frame or the Steam Link app and play them there.

Update, March 11, 4:40PM ET: Updated headline and added details on Valve’s Steam Machine Verified program.

Advertisement

Source link

Continue Reading

Tech

WhatsApp introduces parent-managed accounts for pre-teens

Published

on

WhatsApp

WhatsApp has begun rolling out parent-managed accounts for pre-teens, allowing parents and guardians to decide who can contact them and which groups they can join.

These accounts are restricted to messaging and calling and do not include access to Meta AI, Channels, Status, or location sharing. The child’s messages remain end-to-end encrypted and cannot be viewed by third parties, including parents, who cannot read their chats or listen in on their calls.

Setting up a parent-managed account requires both the parent’s and the child’s devices to be present at the same time. The parent must register and verify the child’s phone number, confirm the child’s age, and scan a QR code on the child’s device to link the two accounts.

The parent can also set a 6-digit PIN, ensuring that only they can access and change message requests, privacy settings, and activity alerts from the managed device.

Advertisement

“The new parental controls and settings are gated by a parent PIN on the managed device. Only parents can access and change privacy settings, ensuring they are empowered to tailor their family’s experience,” the company said.

“All personal conversations remain private and protected with end-to-end encryption, meaning no one—not even WhatsApp—can see or hear them.”

By default, managed accounts can exchange messages only with people already saved in the child’s contacts, and only parents can add the account to groups.

Advertisement

When an unknown contact attempts to reach the child, a context card will appear showing whether the unknown contact shares any groups with the child and which country they are from.

Parents will also receive activity alerts when their kids receive new chat requests, add a new contact, or when groups they’re in add new members.

“Parents will receive notifications about important activity, like when their child receives a message request from an unknown contact. They can customize the notifications they receive about other activities, such as when their child leaves a group,” WhatsApp added.

The child can transition their account to a standard WhatsApp profile when they turn 13, after which they will have full access to WhatsApp features without parental controls.

Advertisement

Meta also introduced dedicated accounts for teens under 16 on Facebook and Messenger in September 2025, one year after rolling out the same feature for Instagram users.

Today, Meta also introduced new WhatsApp anti-scam protections that warn users when behavioral signals suggest a device-linking request may be fraudulent.

Malware is getting smarter. The Red Report 2026 reveals how new threats use math to detect sandboxes and hide in plain sight.

Download our analysis of 1.1 million malicious samples to uncover the top 10 techniques and see if your security stack is blinded.

Source link

Advertisement
Continue Reading

Tech

3D Printing A Harmonic Pin-Ring Gearing Drive

Published

on

Cycloidal drives are a type of speed reducer that are significantly more compact than gearboxes, but they still come with a fair number of components. In comparison, the harmonic pin-ring drive that [Raph] recently came across as used in some TQ electric bicycles manages to significantly reduce the number of parts to just two discs. Naturally he had to 3D model his own version for printing a physical model to play with.

How exactly this pin-ring cycloidal drive works is explained well in the referenced [Pinkbike] article. Traditional cycloidal drives use load pins that help deal with the rather wobbly rotation from the eccentric input, but this makes for bulkier package that’s harder to shrink down. The change here is that the input force is transferred via two teethed discs that are 180° out of sync, thus not only cancelling out the wobble, but also being much more compact.

It appears to be a kind of strain wave gearing, which was first patented in 1957 by C.W. Musser and became famous under the Harmonic Drive name, seeing use by NASA in the Lunar Rover and beyond. Although not new technology by any means, having it get some more well-deserved attention is always worth it. If you want to play with the 3D model yourself, files are available both on GitHub and on MakerWorld.

Advertisement

Source link

Continue Reading

Tech

Tech Moves: Microsoft Research gets a new leader; Amazon head joins AI startup; JPMorgan exec departing

Published

on

Peter Lee, left, and Igor Tsyganskiy. (GeekWire File Photo and LinkedIn Photo)

Igor Tsyganskiy is now executive vice president of Microsoft Research (MSR) as past president Peter Lee steps aside to become president of Microsoft Science, which spans physical, biological and medical fields.

“My new role is designed to reduce my management responsibilities and let me spend as much of my time as possible on technical work,” Lee said on LinkedIn. That will include an initial focus on advances in “AI-enabled virtual patients, populations and labs, and their power to transform biomedical research.”

Lee, who took the helm of MSR in September 2022, was previously a computer science professor at Carnegie Melon University for more than two decades. He thanked Tsyganskiy for “taking on the big job of leading Microsoft Research — I have no doubt that he’ll take the MSR labs up to new heights.”

Tsyganskiy will also continue serving as the tech giant’s global chief information security officer, a role he has held since 2023. In his own LinkedIn post, Tsyganskiy emphasized MSR’s role at the forefront of computing, pointing to advances in AI, deep systems work and scientific discovery that have fed into Microsoft products and academic publications.

The commitment to foundational research is essential to Microsoft’s success, he said, adding “as the pace of innovation accelerates it is equally important to continue driving breakthrough research, and translate these advances into real-world impact.”

Advertisement
Mamtha Banerjee. (LinkedIn Photo)

Mamtha Banerjee is leaving her role as leader of JPMorgan Chase’s Seattle Tech Center. Banerjee, a longtime Seattle tech industry leader, joined the financial services giant in 2022 and took the leadership role last June.

The Seattle Tech Center was established in 2018 to tap into the region’s tech talent pool and by last year had grown to 380 people.

Banerjee was previously at Expedia Group for seven years and serves as a mentor for the University of Washington’s Master of Science in Entrepreneurship program.

Vivian Sun. (LinkedIn Photo)

Vivian Sun is leaving her role as head of automated driving at Amazon after more than two years with the company.

Sun, based in Sunnyvale, Calif., is now vice president of commercial and strategy for robotics company Genesis AI, which is based in Palo Alto and Paris.

A veteran startup builder with roots in AI, robotics and autonomous driving, Sun was featured by Automotive News as one of the “100 Leading Women in the North American Auto Industry.”

Advertisement

Truveta‘s hiring run continues. The Seattle-area health data company named Sapna Prasad as its new VP of research insights. Prasad, who is based in Washington, D.C., joins from Clarify Health Solutions where she held leadership roles for more than six years.

The company recently announced more than a dozen new hires and in January named Dr. Johnathan Lancaster as its president and chief scientific officer.

Bluesky CEO Jay Graber announced Monday that she’s stepping down from her position and moving to a new role as chief innovation officer of the decentralized social network. Read more.

Jessica Nguyen. (LinkedIn Photo)

Jessica Nguyen is now president, chief strategy and legal officer for Sandstone, a New York-based company using AI to support legal work.

Nguyen is based in the Seattle area and and was previously deputy general counsel for AI innovation and trust at DocuSign for nearly two years. Prior to that, she was chief legal officer at Seattle’s Lexion, which was acquired by DocuSign for $165 million in 2024.

Advertisement

Her Pacific Northwest roots include working as the first in-house attorney for Payscale and Avalara, and she had a nearly four-year run at Microsoft on the Office 365 legal team.

Julie Keef, who recently left Seattle’s Redfin as VP of product, has shared her next role. Keef has moved to another company in the real estate space, taking the title of head of consumer product management for New York’s Compass.

“Since my days in NYC, I’ve been admiring Compass from afar. The agents, brand, and bold strategy have always impressed me,” Keef said on LinkedIn.

Nick Boone. (LinkedIn Photo)

Nick Boone is now global head of demand and marketing operations at Scala, a Bellevue-based AI startup founded by Smartsheet CEO Rajeev Singh and former Accolade executive Ardie Sameti. The company last month raised $8.5 million in a seed round.

Boon worked at Accolade for more than eight years, serving as senior director of demand center and marketing operations until the company was acquired by Transcarent last year. He remained with the merged businesses for a brief time.

Advertisement

Scala is building an “operational intelligence platform” for contact centers — the massive customer service operations that companies across healthcare, travel, and financial services rely on to handle millions of interactions.

ProbablyMonsters expanded its executive leadership team with two new hires and a promotion. The video game company is headquartered in Fort Worth, Texas, and has an office in Bellevue, Wash., where founder and CEO Harold Ryan is based.

  • Jonathan Lander was named chief publishing officer. He was previously at Bethesda Softworks and ZeniMax Online Studios.
  • David Reid joins as chief marketing officer and will be located in Bellevue. Reid is a longtime gaming exec who was the founder of Seattle-area startup MetaArcade and more recently ran his own consultancy.
  • Mark Subotnick, who is based in Portland, is now chief product officer after previously serving as head of studios and partnerships. He’s been with the company for more than three years.
Amber Faust. (LinkedIn Photo)

— Biotech startup Nautilus appointed its first sales hire, naming Amber Faust as vice president of sales as it ramps up commercial operations. Faust, who will work remotely, joins the Seattle company after working at biotech businesses including Seer, Olink, SomaScan and others.

“I’m excited to help scale Nautilus’ commercial progress by connecting researchers in pursuit of greater proteomics coverage, detail, and resolution with a platform that can meaningfully expand what’s possible in drug development and beyond,” Faust said in a statement.

Nautilus has built a proteome analysis platform that allows researchers to identify and quantify the thousands of proteins present in biological samples.

Advertisement

Absci, a Vancouver, Wash.-based company that uses AI to develop drugs, named Dr. Ransi Somaratne as chief medical officer, joining from Vertex Pharmaceuticals where he served as senior VP of clinical development. Past roles include leadership positions at BioMarin Pharmaceutical and Amgen.

Absci Chief Innovation Officer Andreas Busch is retiring March 31 and will continue to co-chair the company’s scientific advisory board.

Theo Angelis was appointed to the Washington State Supreme Court. The K&L Gates partner has 25 years of legal experience and has worked extensively in intellectual property and with emerging companies. Angelis is a past president of the Middle Eastern Legal Association of Washington and will be the first Justice of Middle Eastern descent on the state Supreme Court.

Matt Rubright is now CEO of Jam, a startup that addresses bugs in software development. He was previously chief customer officer for the 6-year-old company, joining last April. His past employers include DataGrail, Candidate and Silicon Valley Bank.

Advertisement

Rubright, based in Seattle, succeeds Dani Grant, whom he said is stepping away to recover from a health issue. Grant’s “vision, leadership and tenacity are undeniable,” he added.

Suchitra (Suchi) Mohan is founder of a Sammamish, Wash.-based startup called learntheropes.ai, which she describes as an “AI-native learning app designed to help organizations empower their people with the right learning content, tailored to their learning style, skill-level and their goals, reducing the overwhelm employees feel with a new task.”

Mohan is a serial entrepreneur and worked as a technology architect for Microsoft for nearly a decade in its Bangalore offices. She was most recently co-founder of the AI startup Oikyu.

Theo Michel joined Seattle’s Bayou Energy as senior product engineer. Earlier in his career, Michel was with Micrsoft’s Xbox Live for more than 17 years. The clean energy startup recently named Yoon Loong Wong (Andrew) as chief of staff.

Advertisement

Brian Marrs was promoted to general manager of energy markets for Microsoft, previously serving in a senior director role. He has been with the company for nearly nine years.

Source link

Continue Reading

Tech

Anthropic gives Claude shared context across Microsoft Excel and PowerPoint, enabling reusable workflows in multiple applications

Published

on

Anthropic has upgraded its Claude AI model with new capabilities for Microsoft Excel and PowerPoint, marking a strategic move to expand its enterprise footprint and potentially challenging Microsoft’s newly launched Copilot Cowork — which Claude also partially powers. 

The updated add-ins are available to Mac and Windows users on paid Claude plans starting today, March 11.

Anthropic is also expanding how enterprises can deploy the tools.

Claude for Excel and Claude for PowerPoint can now be accessed either through a Claude account or through an existing LLM gateway routing to Claude models on Amazon Bedrock, Google Cloud Vertex AI or Microsoft Foundry.

Advertisement

That gives enterprises more flexibility to use the add-ins within cloud and compliance setups they may already have in place.

Shared context across Office apps

Starting March 11, paid Claude users on Mac and Windows can access a new beta experience in which Claude for Excel and Claude for PowerPoint share the full context of a user’s conversation with the AI model between the two applications — no need for manually copying and pasting it over.

That means Claude can carry information, instructions and task history between an open spreadsheet and an open presentation in a single continuous session.

For example, Claude can write formulas to extract data from an Excel workbook and immediately apply it to a stylized PowerPoint slide in the same session.

Advertisement

“In practice: a financial analyst can ask Claude to pull comparable company financials from an open workbook, build out a trading comps table in Excel, drop the valuation summary into the pitch deck, and draft the email to the MD—without switching tabs or re-explaining the dataset at each step,” Anthropic said in a press release. 

This builds on Anthropic’s release of a Claude plugin for Excel back in October 2025.

Repeatable workflows inside applications 

A central feature of this launch is Skills, which allows teams to build and save repeatable workflows directly inside the Excel and PowerPoint sidebars.

Rather than re-uploading references or re-prompting instructions, users can save standardized processes—such as specific variance analyses or approved slide templates—as one-click actions available to the entire organization.

Advertisement

That could include workflows for recurring financial analysis, preparing presentations in a preferred house style or running common review steps that would otherwise need to be rewritten as prompts each time.

Anthropic said every Skill, whether personal or organization-wide, will work inside the add-ins the same way MCP connectors do.

“Workflows that previously lived in one person’s head become one-click actions available to the whole organization,” the company said. 

Anthropic distinguishes these Skills from Instructions, which let users set persistent preferences across the add-ins, such as preferred number formatting in Excel or presentation-writing rules in PowerPoint.

Advertisement

Anthropic is also shipping a preloaded starter set of Skills, including:

  • Excel: Auditing models for formula errors, populating DCF and LBO templates, and cleaning messy data ranges.

  • PowerPoint: Building competitive landscape decks and reviewing investment banking materials for narrative alignment.

Similarly, Microsoft’s new Copilot Cowork capability introduced on Monday enables enterprise users to deploy agents to complete tasks across Microsoft applications such as Excel and PowerPoint. 

The software giant openly stated it was built in conjunction with Anthropic, which also released its own stand-alone Claude Cowork application for Mac and Windows earlier this year offering a way for Claude to access, edit, create and move information between files on a user’s computer, autonomously, at the user’s direction.

Previously, even with autonomous tools like the standalone Claude Cowork app, users often had to ask the AI to complete tasks in separate steps for each application. Now, Claude maintains a continuous session that reads live data and writes formulas across both apps simultaneously.

Advertisement

Battle of the enterprise app agents

Ever since the launch of Claude Cowork earlier this year, Anthropic has been making a case to be the chat and productivity platform of choice for enterprises.

Competitors like Google, with its close association with Google Workspace, which includes Gmail and Google Docs, and Microsoft, with its continued leadership in the Office suite, can directly bring AI capabilities to users’ workflows. 

Anthropic did not present the new Skills feature as equivalent to the more autonomous, agentic behavior Microsoft is now emphasizing with its own Copilot Cowork.

But the release does show Anthropic steadily expanding beyond chatbot use cases and into more structured, repeatable work inside the applications many business users already rely on.

Advertisement

Anthropic, through Claude Cowork, Claude Code and the Claude model family, has seeped into many organizations’ systems, using its high performance in coding benchmarks and general knowledge to navigate a computer better and complete knowledge work rapidly, at scale, with high quality.

OpenClaw, the open source AI agent that has taken the developer world by storm, owes much of its existence to Claude Code. 

The result is another sign that the battle over enterprise AI is no longer just about which model performs best on benchmarks. It is increasingly about what AI tools and systems enterprises trust to get real work done across their existing applications, files, and workflows.

Source link

Advertisement
Continue Reading

Tech

Alternative app store AltStore PAL joins the fediverse

Published

on

AltStore PAL, an alternative app store for iOS, made possible by new regulations in markets like the EU and Japan, is integrating with the open social web. The company on Wednesday announced support for the fediverse, the open social web that runs on ActivityPub, which underpins apps like Mastodon, Flipboard, PeerTube, Threads, and others.

Known broadly as the “fediverse,” users on the open social web join independent servers that interoperate with one another, or “federate,” allowing people to communicate, view, and respond to posts with those on other social platforms.

With AltStore PAL’s launch of its own Mastodon server, it’s becoming the first federated app marketplace.

Image Credits:AltStore PAL

The company intends to use its server to give app developers a way to communicate about their apps’ news, updates, and alerts. Because these posts are federated, they can be seen, liked, and interacted with by users across the open social web, including Mastodon, Instagram’s Threads, and even Bluesky — assuming the Bluesky users have bridged their accounts to make them accessible to those in the fediverse.

“So that means, like, if you have a Mastodon account or Threads account, you could follow these accounts. You could follow the source from our Mastodon server, and then, in your timeline, you would see when there was an app update,” AltStore PAL co-founder Riley Testut told TechCrunch last fall, when the company was teasing its plans and announcing its $6 million Series A.

Advertisement

The interactions on these posts will also be shown on the AltStore PAL app itself, as people can sign in with their Mastodon or Bluesky accounts to like the apps, the app updates, and the news alerts within the app marketplace.

Image Credits:AltStore PAL

At launch, a handful of apps are already participating, including Loops, a federated short-form video app; PeerTube, a federated app similar to YouTube; and iPhanpy, a Mastodon client built by indie developer Matt Fantinel.

Some larger apps are planning to launch on this federated marketplace at a later time after changes to Apple’s commission structure are finalized in the E.U. (Due to potential liability issues over payments, these apps are waiting for Apple to replace its Core Technology Fee with its Core Technology Commission, a decision it made after getting pushback from regulators over its prior revamp of its commission system.)

To access the AltStore PAL’s Mastodon server, users can visit fosstodon.org/@altstore. Meanwhile, users can also discover apps and sources on the AltStore PAL’s explore page at explore.alt.store. Alongside this update, the AltStore PAL app has been updated with an iOS 26 Liquid Glass design and icon.

Source link

Advertisement
Continue Reading

Tech

Lyngdorf SB-75 Passive Soundbar Announced: High-End Audio Upgrade for 75″ and 77″ TVs

Published

on

Lyngdorf Audio has introduced the SB-75, a high-end passive soundbar designed to complement 75- and 77-inch televisions, with wider cabinet sizes available by special order. Expected to arrive in the U.S. by Summer 2026, the SB-75 continues the company’s push into luxury home theater solutions alongside its FR-2 loudspeakers, Cue-100 speaker system, and the TDAI-2210 integrated streaming amplifier.

Unlike the $500 all-in-one soundbars stacked by the pallet at big-box retailers, passive soundbars like the SB-75 target a very different audience. These are buyers who already own serious amplification, care about room correction and system matching, and expect a soundbar to deliver the same level of fidelity they demand from their two-channel music systems. In other words, this is for the listener who wants one elegant LCR solution beneath the TV without sacrificing the kind of clarity, dynamics, and scale normally associated with a dedicated speaker system.

That places the SB-75 in a niche but growing category occupied by brands such as Theory Audio Design, which specializes in custom-width passive LCR soundbars designed for 65- to 85-inch displays. Like those systems, Lyngdorf’s approach assumes the use of external amplification, advanced room setup software, and higher-end A/V processors, creating a performance-first alternative to the convenience-driven soundbars that dominate the mainstream market.

SB-75 Overview: Bring Your Own Power

lyngdorf-audio-sb-75-passive-soundbar-gray-angle
Lyngdorf SB-75 passive soundbar in gray

The SB-75 combines two high-end stereo speakers within a single cabinet, giving it the form factor of a soundbar but the design philosophy of a traditional hi-fi loudspeaker. Intended to be wall mounted below a TV or projection screen, the SB-75 is engineered to deliver a wide soundstage and powerful output for both movies and music.

Unlike most soundbars, however, the SB-75 does not include built in amplification. Instead, it uses standard speaker connections, allowing owners to pair it with the external amplifier or amplifiers of their choice, whether that means a high quality two channel amplifier or a pair of monoblocks. This approach provides far greater flexibility for system matching and performance tuning than the typical all in one soundbar design.

Advertisement

High-End Soundbar Design Built Like Real Speakers

lyngdorf-audio-sb-75-passive-soundbar-gray

The Lyngdorf SB-75 may look like a soundbar, but under the hood it’s closer to a pair of serious hi fi loudspeakers housed in a single cabinet. Designed for on wall installation below a TV or projection screen, the SB-75 combines a slim, space saving footprint with the engineering approach Lyngdorf typically applies to its high end speakers.

Inside the rigid MDF cabinet are six drivers arranged in a stereo configuration. The SB-75 uses four 180 mm (6.5 inch) aluminum cone midrange woofers with 35 mm voice coils, die cast baskets, and vented magnets, paired with two 28 mm (1.1 inch) fabric soft dome tweeters. This configuration allows the SB-75 to function as a full range passive stereo speaker capable of delivering both music and movie soundtracks with convincing scale and clarity.

The cabinet itself is shallow but carefully tuned for placement close to a wall, which improves efficiency and overall output. With a rated sensitivity of 92 dB and a nominal impedance of 4 ohms, the SB-75 is designed to deliver high output levels when paired with quality external amplification.

Lyngdorf lists the SB-75 with a frequency response of 57 Hz to 20 kHz (±3 dB), making it capable of substantial bass performance on its own while still integrating easily with a subwoofer for home theater systems. The soundbar measures 168 x 20 x 11 cm (66 x 7.9 x 4.3 inches), or 11.6 cm (4.6 inches) deep when mounted with the included wall bracket, allowing it to sit discreetly below large flat panel displays while delivering the soundstage of a traditional stereo speaker setup.

Designed for Clean Wall Installation

lyngdorf-audio-sb-75-passive-soundbar-lifestyle-front

The Lyngdorf SB-75 is engineered for simple wall mounting and ships with an integrated wall mount bracket for straightforward installation below a TV or projection screen. Its shallow cabinet profile makes placement easy while maintaining a low visual impact in both dedicated home theaters and everyday living spaces.

Finished in an extra matte black, the SB-75 is designed to visually disappear in dark cinema rooms or when placed behind an acoustically transparent projection screen. For more traditional media rooms and living rooms, the speaker features rounded cabinet edges and an elegant midnight gray fabric grille that helps it blend seamlessly with modern interior design.

Advertisement
Advertisement. Scroll to continue reading.

Pro Tip: The SB-75 is sized to match 75 inch and 77 inch televisions, creating a clean, proportional look beneath large displays. Lyngdorf also offers custom cabinet widths by special order, allowing integrators and homeowners to match the soundbar precisely to other screen sizes.

Steinway Lyngdorf Model S

Steinway Lyngdorf Model S Soundbar Under TV Lifestyle
Steinway Lyngdorf Model S Soundbar

In addition to the SB-75, Lyndorf Audio’s sister company, Steinway Lyngdorf, offers an integrated powered soundbar system, the Steinway & Sons Model S (originally released in 2024).

Instead of a passive design, the Model S incorporates a high-end speaker speaker assembly, but also provides Room Perfect technology for performance optimization in different room acoustic properties and size, and also includes an externally connected Steinway Lyngdorf amplification system.

Comparison

Lyngdorf SB-75 Steinway & Sons Model S
Product Type Passive Soundbar Powered Soundbar
Price $5,000 $17,000
Description 2 x 2-way speaker (stereo soundbar) Full-range on-wall loudspeaker
Enclosure MDF construction, closed cabinet Aluminum front and back, MDF frame
Tweeter 2 x 28mm (1.1″) Fabric soft dome 3 x Air Motion Transformer, Kapton foil, neodymium magnet
Midrange 4 x 180 mm (6.5″), 35mm voice coil, cone material: aluminum, die-cast basket, vented magnet  3 x 5.25″ anodized aluminum cone, neodymium magnet 
Woofer 2 x 10″ aluminum cone, 4-layer voice coil, neodymium magnet
Nominal Impedance 4 Ohms N/A
Frequency Response (-3dB) 57 – 20,000 Hz 40 Hz – 20 kHz
Power Handling 250 Watts 1,600 Watts
Amplification External A1 or A2 amplifiers, 4 x 400 watt  (left channel, right channel, center channel, dual woofers)
Max SPL@1m 117 dB 113 dB
Sensitivity (2.83V/1m) 92 dB N/A
Crossover 2,200 Hz Woofer/midrange: Digital, stored in processor (200 Hz, 4th order) 
Advertisement

Midrange/tweeter: (2,500 Hz, 4th order)

Connections Binding posts (2 pairs; one for left and one for right channel) 2 x Neutrik speakON NL4
Placement On-wall with supplied wall-mount bracket On-wall
Dimension (WHD) 168 x 20 x 11 cm (11.6 cm with wall mount bracket)

66 x 7.9 x 4.3 inches (4.6 inches with wall mount bracket)

Custom Width available by special order

Advertisement
189 x 30 x 10.4 cm

74.4 x 11.8 x 4.1 inches (incl. base plinth)

Weight 28 kg / 62 lbs 48 kg / 106 lbs
Finish Ultra matte black, detachable fabric grille in midnight grey Matte black and high-gloss black with gold accents. Customized lacquers and finishes are available on request.  
lyngdorf-audio-sb-75-passive-soundbar-front
Lyngdorf SB-75

The Bottom Line 

The Lyngdorf SB-75 isn’t meant for shoppers comparing $300 soundbars at the local electronics store. It’s aimed squarely at enthusiasts and custom installers who want the clean look of a soundbar but refuse to give up the performance of real loudspeakers. By using a fully passive design with external amplification, Lyngdorf offers far more flexibility and system matching than the typical all in one soundbar.

For listeners who value both music fidelity and cinematic impact and already own high quality amplification or plan to build a serious home theater system, the SB-75 delivers something unique: the sound of high end stereo speakers in a single, elegant cabinet designed to sit seamlessly beneath a large screen.

lyngdorf-audio-sb-75-grill-logo-closeup
Lyngdorf logo on SB-75 grille

Price & Availability

The Lyngdorf SB-75 is expected to become available by Summer 2026 via Authorized Dealers for an approximate price of $5,000 USD (that does not include required amplification).

Note: The Steinway Lyngdorf Model S is currently available at Authorized Dealers for $17,000.

Advertisement

Source link

Continue Reading

Tech

Honor of Kings Is Finally Available in India (Free Skins, Events, and Esports)

Published

on

Tencent’s TiMi Studio and Level Infinite have officially launched Honor of Kings in India. The mobile MOBA game, often described as the world’s most-played MOBA, is now available to download on both Android and iOS devices starting today. The game follows a “Free to Play, Fair to Win” philosophy, meaning players can progress and compete based on skill rather than paid advantages. With the India launch, Honor of Kings is introducing localized content, community events, and esports opportunities aimed specifically at Indian players.

Classic 5v5 MOBA Gameplay Comes to Indian Players

Honor of kings character holding a sword

At its core, Honor of Kings brings traditional 5v5 MOBA gameplay to mobile devices. Matches take place in Hero’s Gorge, a battlefield where teams compete to destroy the enemy base through strategy, teamwork, and hero abilities. Players can choose from a wide roster of heroes with different roles and abilities. The game also features quick matchmaking, ranked battles, and localized voiceovers designed to improve the experience for Indian users.

Interestingly, Dean Huang, Producer of Honor of Kings, said the company will invest ₹100 million in India to support creators and the esports ecosystem, which is amazing news for the eSports community.

Launch Events Offer Free Heroes and Skins

Character in honor of kings game

To celebrate the India launch, Honor of Kings is running several limited-time in-game events that reward players with free items. The “Sign In for a Free Epic Skin” event runs until April 11, allowing players to unlock rewards by logging in regularly. These include heroes such as Mai Shiranui, Ziya, Ukyo Tachibana, Haya, Charlotte, and Augran, along with the Epic Skin Feline Whisperer for Mai Shiranui.

Beyond that, another event, “Game Paltega: Desi Legends United,” runs until March 24. This influencer-led challenge allows players to earn redemption coins that can be exchanged for rewards such as avatar frames, creator voice packs, and stickers featuring popular gaming creators like Scout, Mortal, and Kaash Plays. Players can also participate in the “Draw Daily to Get Gift Cards” event between March 13 and March 21, and the game plans to introduce a new hero inspired by Indian culture as well.

Esports Plans for India

We’ve been to Tencent’s eSports events, and they have always been amazing. Keeping that spirit in mind, the company is building an esports pipeline for Indian players. In fact, two Indian teams will get slots in the KWC at EWC26, the global Honor of Kings tournament pathway. Additionally, the KINGS’ Arise: India City Tour will host offline finals in multiple cities:

Advertisement
  • Bengaluru — March 22
  • Mumbai — March 29
  • Delhi — April 5

Each event will feature a ₹100,000 prize pool.

Source link

Continue Reading

Tech

Microsoft’s brief in Anthropic case shows new alliance and willingness to challenge Trump administration

Published

on

Microsoft said the action against Anthropic imposes “substantial and wide-ranging costs and risks.” (GeekWire File Photo / Todd Bishop)

A brief filed by Microsoft in Anthropic’s lawsuit against the U.S. Department of War shows the deepening ties between the two companies, and Microsoft’s willingness to take on the federal government at key moments in its history.

Microsoft on Tuesday urged a federal judge in San Francisco to temporarily block the Pentagon’s designation of Anthropic as a supply chain risk, arguing that immediate enforcement would hurt Microsoft and other government contractors that depend on Anthropic’s technology.

The government’s designation imposes “substantial and wide-ranging costs and risks” on companies that use Anthropic’s models “as a foundational layer of their own products and services, which they provide to the U.S. military,” Microsoft said in the filing.

The New York Times DealBook called Microsoft’s brief “a remarkable act” and “a momentous decision” for a company that is one of the largest government contractors in America, noting that it stands out in a period when corporate America’s unwritten rule has been to avoid picking fights with the White House.

It came a day after Microsoft launched Copilot Cowork, a new AI product built on Anthropic’s Claude models, and four months after Microsoft committed to invest up to $5 billion in the startup in a deal that includes  Anthropic spending at least $30 billion on Microsoft Azure.

Advertisement

Amazon, which has invested $8 billion in Anthropic, has not publicly weighed in on the lawsuit or the supply chain risk designation. We’ve contacted the company for comment.

Microsoft hasn’t shied away from fighting with Washington, D.C., at key moments in its history, ranging from its landmark antitrust battle with the Justice Department in the late 1990s to its Supreme Court fight against the Trump administration over DACA immigration protections. 

The Redmond-based company has built one of the deepest government-relations operations in tech, led by President and Vice Chair Brad Smith, a former D.C. lawyer whom the New York Times once called “a de facto ambassador for the technology industry at large.”

Anthropic sued the Department of War on Monday over the designation, which is historically reserved for foreign adversaries. It followed the collapse of contract negotiations in which Anthropic refused to drop two guardrails on its AI models: no use for fully autonomous weapons and no use for mass domestic surveillance of Americans.

Advertisement

President Trump separately directed all federal agencies to stop using Anthropic’s technology.

OpenAI, meanwhile, moved quickly to fill the gap left by Anthropic, announcing its own Pentagon deal on the same day the designation came down. CEO Sam Altman later acknowledged the timing looked “opportunistic and sloppy.” Thirty-seven engineers and researchers from OpenAI and Google, including Google chief scientist Jeff Dean, separately filed their own amicus brief in support of Anthropic.

In its amicus brief, Microsoft said AI should not be used “to conduct domestic mass surveillance or put the country in a position where autonomous machines could independently start a war,” aligning itself with Anthropic’s position on the two sticking points in the negotiations.

Microsoft also flagged a double-standard in the government approach: the Pentagon gave itself six months to transition off Anthropic’s models but made the designation effective immediately for contractors. Without a restraining order, Microsoft warned, it and other companies would have to “act immediately to alter existing product and contract configurations” for the military.

Advertisement

Source link

Continue Reading

Tech

Valve defends loot boxes in response to New York’s lawsuit

Published

on

It must be 2017 because loot boxes are back in the news again. Two weeks after New York’s attorney general sued Valve over its use of the gimmick, the company has responded. In short, the Steam maker essentially said, “See you in court.”

New York’s lawsuit accuses Valve of promoting illegal gambling through its games. AG Letitia James called the loot boxes found in titles like Counter-Strike 2, Team Fortress 2 and Dota 2 “addictive, harmful and illegal.” The state seeks to “permanently stop Valve from continuing to promote illegal gambling in its games” and pay relevant fines.

In its defense posted on Thursday, Valve likened its mystery boxes to kids buying packs of physical trading cards. “Players don’t have to open mystery boxes to play Valve games,” the company wrote. “In fact, most of you don’t open any boxes at all and just play the games — because the items in the boxes are purely cosmetic, there is no disadvantage to a player not spending money.”

That last point, while applicable within the game itself, isn’t quite that cut and dry once you zoom out beyond that. As James pointed out, players can trade the cosmetic items they win from loot boxes on Steam’s marketplace or sell them on third-party marketplaces. Rarer ones can sometimes fetch lucrative sums.

Advertisement
CS2 gun skin listed for $20,000 on a marketplace

A CS2 gun skin listed for $20,000 on DMarket (DMarket)

Here, too, Valve defended the profitable practice by rolling out the trading card comparison. “We think the transferability of a digital game item is good for consumers — it gives a user the ability to sell or trade an old or unwanted item for something else, in the same way an owner can sell or trade a tangible item like a Pokémon or baseball card,” the company wrote. “NYAG proposes to take away users’ ability to transfer their digital items from Valve games. Transferability is a right we believe should not be taken away, and we refuse to do that.”

Valve is also facing a new class-action lawsuit over its loot boxes.

Some of Valve’s points land a bit more than its righteous defense of a gaming gimmick that, well, isn’t exactly beloved. The company accused the NYAG of proposing that Valve collect additional user information to prevent VPN use. In addition, the state allegedly “demanded that Valve collect more personal data about our users to do additional age verification.” Privacy experts have been sounding the alarm about the recent push for online age verification.

Valve also addressed James’s erroneous and outdated statement that video games encourage real-world violence. “Those extraneous comments are a distraction and a mischaracterization we’ve all heard before,” the company wrote. “Numerous studies throughout the years have concluded there is no link between media (movies, TV, books, comics, music and games) and real world violence. Indeed, many studies highlight the beneficial impact of games to users.”

Advertisement

The company says that, while it may have been cheaper to settle the suit, it deemed the NYAG’s demands user-hostile. “Ultimately, a court will decide whose position — ours or NYAG’s — is correct. In the meantime, we wanted to make sure you were aware of the potential impact to users in New York and elsewhere.”

Source link

Continue Reading

Trending

Copyright © 2025