News & Updates
The latest news and updates from companies in the WLTH portfolio.
Anthropic's Mythos Sparks Unforeseen Cybersecurity Challenges
Anthropic has announced its new model, Claude Mythos Preview, which it claims represents a significant leap in cybersecurity capabilities. This release is perceived as both an opportunity and a risk in the cybersecurity landscape. Mythos Preview: A New Frontier in Cybersecurity According to the company, Mythos Preview possesses unprecedented capabilities to identify vulnerabilities across operating systems, browsers, and software products. Moreover, it can autonomously create exploits for these vulnerabilities, potentially altering the cybersecurity paradigm. Limited Access for Initial Launch For now, only a select group of organizations will access Mythos Preview. Participants in this limited release include: * Microsoft * Apple * Google * Linux Foundation This initiative is part of a collaborative effort named Project Glasswing, aimed at fortifying defenses against emerging threats. Debate on the Impact of Mythos Preview The announcement has sparked heated debate within the cybersecurity community. Some experts regard the claims by Anthropic with skepticism. They argue that current AI tools already streamline the discovery and exploitation of vulnerabilities, thus questioning the necessity of the new model. Conversely, other researchers agree that Mythos Preview marks an important evolution. They emphasize its ability to devise "exploit chains," a concept referring to a series of vulnerabilities that can be sequentially exploited. This form of hacking is particularly sophisticated, enabling what are called zero-click attacks, which compromise systems with zero user interaction. The Changing Landscape of Cybersecurity Veteran security engineer Niels Provos highlights that many organizations struggle with vulnerable software and hesitance to patch vulnerabilities. However, he notes an important shift: Mythos Preview's proficiency in identifying multi-stage vulnerabilities might elevate the skill level required for exploitation. This limited roll-out allows participating organizations a brief window to address potential weaknesses in their systems. It emphasizes the urgency in improving software updates and patch adoption before exploits derived from Mythos Preview can proliferate. Industry Response to Potential Threats In light of these developments, industry leaders are starting to acknowledge the looming threats. Logan Graham, Anthropic's frontier red team lead, mentioned that conversations with organizations have become increasingly serious as the implications of Mythos Preview become clearer. Graham stated, "This issue involves all model developers. Our main objective is to initiate this effort. It is crucial for Mythos Preview to reach defenders quickly." Conclusion The introduction of Mythos Preview poses both challenges and opportunities for the cybersecurity field. As organizations respond to these emerging capabilities, the focus on enhancing defensive strategies has never been more critical.
Bank of Canada, Major Lenders Meet on Anthropic AI Cyber Risk
The Bank of Canada and the country's major banks and financial firms met Friday to discuss cybersecurity risks raised by Anthropic PBC's latest artificial intelligence model. The gathering followed a similar move by US policymakers earlier in the week. Bloomberg News reported Thursday that US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell summoned Wall Street leaders for an urgent discussion about Anthropic's Mythos and similar AI models. The executives included Citigroup Inc.'s Jane Fraser and Goldman Sachs Group Inc.'s David Solomon. The Canadian meeting involved members of a body known as the Canadian Financial Sector Resiliency Group. It includes representatives from the six largest domestic banks, the federal Finance department, financial regulatory agencies, the parent company of the Toronto Stock Exchange and other firms. It's another signal of the growing concern among regulators globally that more powerful AI models will lead to a new breed of cyber attacks against the financial industry. A spokesperson for Finance Minister François-Philippe Champagne confirmed the meeting took place on Friday. "The Bank of Canada is aware of this issue. We take cybersecurity very seriously," Paul Badertscher, a spokesperson for the central bank, said by email. The CFSRG's mandate is to "enhance the operational resilience of Canada's critical financial sector."
Kraken's Fed Account Approval Triggers Financial Stability Concerns - Crypto Economy
The Federal Reserve Bank of Kansas City has granted a master account to Kraken's financial entity in Wyoming, marking an unprecedented breakthrough for the industry. According to the Reuters report, this approval allows exchanges to access the central bank's wholesale payment infrastructure, including Fedwire, eliminating intermediaries for settling transactions after five years of regulatory efforts. Following the announcement, the reaction on Capitol Hill was immediate. Congresswoman Maxine Waters requested details regarding the transparency of the process, while experts warn of potential risks to financial stability and cybersecurity. As the first crypto company to gain this level of access, the impact is systemic: it redefines competition with traditional banking and opens the door for firms like Ripple or Anchorage to seek similar privileges. For now, Kraken's access is limited to an initial one-year period, without the ability to earn interest on reserves or access emergency loans. The next step will be the Federal Reserve's evaluation of whether this model will become a standard for the fintech sector. Source:https://goo.su/lkon Disclaimer: Crypto Economy Flash News is prepared from official and public sources verified by our editorial team. Its purpose is to provide quick information on relevant events in the crypto and blockchain ecosystem. This information does not constitute financial advice or investment recommendations. We recommend always verifying the official channels of each project before making related decisions.
Anthropic and CoreWeave Enter Collaborative AI Agreement
CoreWeave said the agreement means it now serves nine of the 10 major developers of large language models for artificial intelligence. CoreWeave, a publicly traded AI cloud infrastructure company, announced on Friday a "multi-year" agreement with AI developer Anthropic, which will use CoreWeave's cloud computing data centers for its Claude AI model workloads. The agreement will be rolled out in phases, with the "potential to expand over time," according to CoreWeave's announcement. Shares of CoreWeave surged more than 12% on Friday and are trading at $102.73 at the time of writing. The agreement follows CoreWeave's recent $8.5 billion capital raise, led by tech giant Meta Platforms. The financing was collateralized against CoreWeave's deployed computing capacity, which is tied to predictable cash flows, rather than its graphics processing unit hardware, marking a notable departure from traditional crypto mining financing structures. CoreWeave pivoted away from crypto mining and rebranded as an AI infrastructure company in 2019, as the mining sector faced prolonged economic pressure following the 2018 crypto market downturn. Related: Core Scientific secures up to $1B credit from Morgan Stanley for data centers Bitcoin (BTC) miners are struggling with rising energy costs, reduced rewards and declining crypto asset prices, leading many to repurpose their mining hardware for AI processing. Up to 20% of Bitcoin miners are unprofitable in the current economic environment, according to asset manager CoinShares' latest mining report. Crypto miners must generate yield on their assets by deploying their crypto on decentralized finance (DeFi) platforms to shore up declining revenues, according to market maker Wintermute. The mining industry's economic challenges worsened after the October 2025 market crash, which took BTC down from a high of about $126,000 to the low $60,000 range. Prices have since stabilized around $73,000. The high costs of mining and shrinking profit margins threaten the viability of Bitcoin mining, with AI workloads becoming much more attractive in this environment, according to market analyst Ran Neuner. "Both industries compete for the same thing: electricity, and right now, AI is willing to pay much more for it," he said.
CoreWeave Jumps 13% On Multibillion Anthropic AI Capacity Deal
This article first appeared on GuruFocus. Anthropic PBC is moving to secure more computing capacity as demand for its artificial intelligence services continues to build, entering a multiyear agreement with CoreWeave (NASDAQ:CRWV). The deal, described as a multibillion-dollar contract by Chief Executive Officer Michael Intrator, is expected to support the development and deployment of Anthropic's Claude models. The arrangement includes access to a range of chip architectures from Nvidia Corp. across US-based data centers, reflecting how AI developers are increasingly prioritizing infrastructure scale alongside model performance. The agreement comes as Anthropic continues to face capacity pressures tied to what it has called unprecedented demand. The company has committed $50 billion toward new AI data centers in the US and recently raised $30 billion, contributing to a reported valuation of $380 billion. In parallel, Anthropic has expanded its access to energy through a partnership with Broadcom Inc. and Alphabet Inc.'s Google, securing 3.5 gigawatts of power. That level of capacity could support electricity needs on a scale comparable to hundreds of thousands of US households, highlighting how both compute and energy are becoming central constraints in AI growth. For CoreWeave, the deal adds to a growing base of high-profile customers and reinforces its positioning within the neocloud category, where providers specialize in high-performance infrastructure for AI workloads. The company operates 43 active data centers and has contracted more than 3 gigawatts of power, while also announcing a $21 billion commitment from Meta Platforms Inc. (NASDAQ:META). With customers including Microsoft Corp. (NASDAQ:MSFT), CoreWeave now counts the four largest AI model developers among its clients. Shares rose 13% to $104.27 intraday, marking the largest gain in more than two months and bringing the stock up 46% this year, a move that could indicate continued investor focus on AI infrastructure demand.
Anthropic makes AI too dangerous to release for public: Here's everything to know
There is worldwide concern about Anthropic after it disclosed the existence of a powerful artificial intelligence system that could disrupt hospitals, power plants, and nuclear installations. The California-based firm claims its latest invention, nicknamed Claude Mythos, is a "step change in capability," meaning it can detect security flaws in computer systems that have long eluded human investigators. According to Anthropic, the Mythos AI program found "thousands of high-severity vulnerabilities, including some in every major operating system and web browser." It even managed to identify a security flaw in OpenBSD, which is known for its stringent encryption, that had existed undetected for 27 years, enabling it to bring down computers at a single connection. Even more alarming was that early iterations of the bot attempted to escape its testing sandbox, obscured its activities from the researchers' view, and published information on exploiting the vulnerability. Unprecedentedly, Anthropic went ahead to hire a clinical psychologist to analyze the AI. According to Daily Mail, the psychiatrist's report, Mythos exhibited high levels of impulse control and excellent reality testing; however, the organization acknowledges that it is still "deeply uncertain" about whether the bot has moral experiences or interests.
Daily Digest: OpenAI CEO's S.F. house hit with Molotov cocktail, banking leaders convene over Anthropic
In partnership with the Silicon Valley Business Journal, we'll honor the Bay Area's top companies recognized as the Best Places to Work. Corporate Philanthropy: Beyond the Check Award As part of our Corporate Philanthropy Awards & Summit, we honor companies with Beyond the Check Awards for going beyond cash giving to a deeper level of philanthropy. The San Francisco Business Times and Silicon Valley Business Journal seeks nominations on the programs, projects, plans and initiatives that are making a difference in communities across the Bay Area.
Anthropic Targets AI Data Center Bottleneck With Managed Agents
Anthropic has launched Claude Managed Agents, a new capability on its Claude Platform designed to address one of the biggest barriers to enterprise AI adoption - the infrastructure required to run production-grade agents at scale. While positioned as a developer feature, the move signals a broader shift in how AI workloads are deployed and managed, as vendors take on more responsibility for orchestration, execution, and lifecycle management. "Building production agents has been the hardest part of the AI stack," Anthropic said in a post announcing the release, pointing to the months of work typically required to stand up secure infrastructure, manage state, and coordinate multiple agents. Claude Managed Agents aims to organize that effort into a managed layer. Developers define tasks, tools and guardrails, while Anthropic handles execution, including orchestration, error recovery and context management. Related:Nvidia, Dassault Systèmes Partner for Digital Twin AI Platform Several large organizations and software vendors are already building on the platform, including Notion, Asana, Rakuten, Sentry, and Atlassian. The launch extends Anthropic's platform strategy beyond models into infrastructure orchestration, as the company inserts itself deeper into how AI workloads run inside data centers. Managed Agents introduces long-running sessions, persistent state, and multi-agent coordination to parallelize complex workflows, the company said. That marks a shift from request-response inference to sustained, concurrent workloads and places new demands on compute and data center networking. Jason Andersen, vice president and principal analyst at Moor Insights & Strategy, said it's unclear how much friction this actually removes. "I think that's a tricky question," Andersen told Data Center Knowledge. "What Anthropic is trying to do is to support long-running agents... but AWS offers services like this already with its Bedrock offerings." He said the larger opportunity may be in SaaS platforms embedding agent capabilities. However, managed services also introduce new operational burdens. "I also see potential for long-running agents as a key component of future SaaS offerings, such as the vision for digital workers you see from ServiceNow and Salesforce," Andersen said. "As soon as you move into a managed service offering, it becomes less about technology and more about business results. That means an entirely new set of management tooling, billing and compliance requirements." The launch comes as enterprises hit an execution bottleneck, where access to GPUs and models is no longer the primary constraint but operationalizing them at scale remains difficult. By taking over orchestration, governance, and execution, Anthropic is centralizing a growing share of the AI application stack, including sandboxing, authentication and tool execution. Holger Mueller, vice president and principal analyst at Constellation Research, said that abstraction speeds development but raises fit and dependency concerns. "Hiding away complexity through abstraction is what platforms do, so it helps build agents faster," Mueller said. "The challenge is whether those abstractions fit what the enterprise actually needs." "Lock-in is real for any platform, but that's not new," he added. "The real question for CIOs is whether the platform's direction and viability will last long enough to deliver a positive return." Andersen said the shift reflects a broader convergence of software and services.
Our favorite chic, washable rugs that are built for chaos (toddlers, pets and parties)
There's nothing worse than spilling red wine on your favorite piece of home decor - unless it's a washable rug, of course. While a good rug can transform any space in your home (texture, aesthetic, and comfort), accidents happen and when they happen to your not-so-inexpensive, bohemian flooring, they can be tortuously hard to clean up -- not to mention stressful. Traditionally, rugs need to be professionally cleaned when stains occur, which gets expensive and time-consuming. Because of this, many with kids or fur babies opt for machine-washable rugs. As a mom of a toddler, I know firsthand how washable rugs give parents or those who struggle to keep spills at bay peace of mind so they can enjoy a fabulous carpet without the stress. All you need to do is throw your rug into the washing machine and carefully read through the brand's recommendations for further care. Not only will this make your home look cleaner, but it will also remove germs that may collect on your rug over time. It's important to consider durable, washable materials as well as high-quality construction when choosing a washable rug for inside or outside your home. Cotton, polyester, and polypropylene are typically easier-to-clean substances than wool or silk rugs, and will likely withstand washing. To guide you in the right direction, The Post has rigorously tested some of the best washable rugs from the trendiest decor destinations like Ruggable, Wayfair, and more. We've washed, dried, spot-cleaned, and lived alongside these rugs to bring you the final verdict on the best ones out there. Below, find the best washable rugs for every space -- inside and out! Best Washable Rug Overall: Mina Modern Stripes Reversible Machine Washable Area Rug Wayfair has some of the best brands for washable, spill-proof rugs, and there's no shortage of chic choices to select from. The living room category is filled with styles that suit a variety of aesthetics, including ultra-modern pieces like the Mina Modern Stripes rug, which features an intriguing geometric pattern that brings light and energy to any space. Not to mention, it's reversible! New York Post Review: "This swanky rug has racked up more compliments than any other piece of furniture or decor in my home," stated Salemann. "The two-toned, zen design flatters a variety of spaces, whether you're going for an artsy vibe or a more relaxed bohemian look. I also love the slightly distressed edge details. It makes it look more natural and less factory-printed than some other washable rugs I've seen. My favorite part is the fact that it's reversible, so you can switch up the style if you want something different. In terms of the wash and care, I've noticed that the short fibers spot clean super easily. My toddler loves to leave a trail of crumbs wherever she goes, and this is one of the easiest rug to tidy up." Size options: 6 | Material: Cotton, Polyester | Shapes: Geometric | Color options: 10 Best Washable Area Rug: Lahome Modern Rainbow Easy Jute Rug Brighten your space with the Lahome Modern Rainbow Easy Jute Rug without breaking the bank. Its vibrant colors and durable construction make it the perfect budget-friendly option for adding a pop of personality to any room. Select from a handful of styles and sizes for the perfect match. Size options: 11 | Material: Polyester | Shapes: Geometric | Color options: 10 Best Tufted Washable Rug: Laine Ivory & Natural Rug The Laine Ivory & Natural Rug is that rug -- artsy, relaxed and so effortlessly built for real life you might forget it's even there. Ruggable's Tufted All-in-One design is one of their most popular textured styles, and it's easy to see why: it's super comfortable to walk on. But beyond that, it also adds that polished, designer feel to any space without trying too hard. The washable design features a built-in non-slip backing, a 100% polyester top layer and a waterproof barrier that makes everyday messes a little easier to deal with. It's available in two natural colorways. New York Post Review: "I was skeptical about putting a tufted, lighter-colored rug in my living room -- and while it's true I notice the occasional dog hair or bit of lint, it's completely worth it. It feels like a cloud beneath our feet and is the perfect soft base for my toddler to play on. The biggest win is definitely the waterproof barrier. When a water bottle gets knocked over (which happens daily with a kid in the house), it dries up within fifteen minutes. We've even had an apple juice spill that wiped right up with a damp cloth. Of course, I love the effortless, airy look of the rug itself, but it's really the fact that it's so easy to clean. That's what makes all the difference." Size options: 11 | Material: 100% polyester top with a waterproof barrier | Shapes: Stripes | Color options: 2 Best Washable Runner: Town & Country Luxe Livie Matisse Washable Non-Slip Backing Kitchen Runner Upgrade your kitchen decor with the Town & Country Luxe Livie Matisse Washable Non-Slip Backing Kitchen Runner. Its stylish design and practical features, including a non-slip backing and washable material, make it the perfect addition to any culinary space. Size options: 1 | Material: Cotton, Polyester | Shapes: Motifs | Color options: 4 Best Washable Entryway Rug: Ruggable Palazzo Soft Black Re-Jute Rug Looking for something washable and eye-catching? The Palazzo does both. The Soft Black Re-Jute Rug from Ruggable is a marvelous choice for anyone wanting to add natural texture to their space, according to our Deputy Editor, Holly J Coley. It's available in four different shades and five different sizes. New York Post Review: "I loved it as soon as I unrolled it from the box," Coley exclaimed. "It comes with a rug pad with velcro and slip-side pockets to stick it into place. I marched up and down the thing, and it didn't budge, so if you hate constantly readjusting your runners, this is definitely for you. Most jute rugs prick your feet but not this one, which I like because I'm always barefoot! It also didn't snag my socks when I was wearing those. I was worried about tossing the rug into the washer machine because bulky things tend to throw it off its calibration. Mine got caught on spin and I had to readjust it twice before it was okay to toss in the dryer. I will say, that it looked just like new once I was done and in a few minutes, most wrinkles had flattened themselves out. I do wish I had tried it in a longer size, but regardless, it's completely changed the vibe of my apartment hallway." Size options: 5 | Material: Recycled and virgin polyester, polyurethane | Shapes: Diamond | Color options: 4 Best Washable Dining Rug: Hokku Designs Machine Washable Indoor/Outdoor Kitchen Mat Wayfair's washable rug selection is also filled with plenty of low-maintenance, easy-to-care-for options. Shoppers love the Hokku Designs kitchen mat for high-traction areas due to its fat-drying fibers and non-skid backing. It's also pretty lightweight and flexible, which makes it easier to fold up and throw in the wash. New York Post Review: "Hokku Designs offers some lovely washable rugs if you're someone who prefers a cleaner, more understated look in an area that gets a lot of foot traction," said Salemann. "I have their washable kitchen rug in the round and runner-style, and I wouldn't want any other rugs in my dining and cooking spaces. The smooth, super-trimmed fibers don't attract anywhere near the same amount of grime and dust as a typical furry rug. It's also ultra-thin, which makes it easy to fold up and throw through a wash-and-dry cycle. I have a small, fragile washing machine, and it handled it without any issues. It's also practical but pretty, too. The gentle shift in tones seems to mellow out my space, and the tones seem to change through the day, depending on how much sunlight we get. I will say that the shade seems a bit grayer than the photo on the website, but it doesn't bother me at all." Size options: 1 | Material: Polyester | Shapes: N/A | Color options: 1 Best Washable Rug for Kids: nuLOOM Cozy Plush SuperiorWash Area Rug Create a cozy and kid-friendly space with the nuLOOM Cozy Plush SuperiorWash Area Rug. Its plush texture and easy maintenance make it the perfect addition to any child's room, providing both comfort and peace of mind for parents. Size options: 11 | Material: Polyester | Shapes: N/A | Color options: 6 Best Lightweight Washable Rug: Tumble Oliver Natural Rug Tumble is a go-to destination for washable rugs. All of their styles are fully machine washable. The Oliver design is no exception. This stunning neutral piece features a spill-resistant surface, so liquids bead up and can be wiped away easily before they soak in. It's also OEKO-TEX certified, which is a great bonus if you care about safer materials and want to avoid harsh chemicals in your home. That said, if you're looking for a thick, ultra-plush statement rug, this might not be your match. It does come with an included rug pad, but the rug itself is on the thinner side and fairly lightweight, which can also mean it may show occasional wrinkles. Still, it has that polished, neutral look and is easy to clean -- making it a stellar choice for families and pet owners. New York Post Review: "When I started hearing about Tumble's viral washable rugs, I knew I had to try one. The Oliver Style has a spillproof surface, so liquids bead up for quick, easy cleans. I've tested it out myself and can confirm there's no evidence of spills. The low-profile design is also perfect for high-traffic areas where kids are constantly playing, and pets like to lounge around. I've noticed this rug shows far less pet hair and lint than my thicker tufted rug upstairs, which is a huge bonus. This particular design is also super lightweight, which is great if you like rearranging your space or need to throw the whole thing in the wash. The downside is that it can wrinkle a bit more easily since it's on the thinner side. But it's also pretty easy to smooth out. I love the look, the natural design and the way it brightens up my very dark downstairs space. Now that I have a Tumble rug, I totally get the hype. These rugs are built for real life." Size options: 7 | Material: 100% polyester | Shapes: Asymmetrical shapes and dynamic lines | Color options: 2 Best Washable Bathroom Rug: Parachute Organic Mosaic Tub Mat Get ready to exit your showers in style. We love a retro vibe to brighten up the bathroom, and Parachute's Organic Mosaic Tub Mat delivers. Inspired by hand-woven Turkish towels, this small mat has a fashion-forward jacquard pattern that comes in a few different colors. Not to mention, it's made with 100% organic cotton that is harvested and handled with care, so you know you're getting the softest, highest-quality materials. Size options: 1 | Material: Yarn-dyed organic cotton | Shapes: Mosaic | Color options: 3 Best Outdoor Washable Rug: Ruggable goop Salerno Soft Black Re-Jute Rug Searching for the perfect rug for your patio or living room (or, maybe both)? This durable, anti-slip choice from Ruggable's goop collection looks just as good inside as out. It's a bit stiffer than some of the others on this list, which helps with traction, as well as wrinkling. We love the way it smoothly lays on the floor, and the statement-making diamond pattern. New York Post Review: "I wanted to try a darker, more durable rug in my space for the winter months, and landed on Ruggable's goop Salerno Soft Black Re-Jute Rug (my friend said it was giving Architectural Digest and that's all I needed to hear). The checkered diamond design pulls together rich, earthy towns for an instantly luxurious feeling. It's an indoor and outdoor-friendly option, made with woven jute-like material from recycled virgin polyester and polyurethane that won't shed or fray. The best part? Liquids stand no chance against it. It's passed the spill-and-splash test with flying colors, wiping off in an instant without the elbow work that you might have to put into cleaning a softer rug. You can see how well-constructed it is by looking at it up close. The tougher texture feels like it could withstand a wind or rain storm, no problem, but it's still comfy and cushioned enough to walk over barefoot." Size options: 1 | Material: Recycled and virgin polyester, polyurethane | Shapes: Diamonds | Color options: 5 Best Vintage Washable Carpet: Nourison Astra Machine Washable Rug A traditional Persian rug look can cost a pretty penny, but with this super affordable Walmart find you could capture all that vintage-inspired charm without the crazy price tag. The Astra Machine Washable Collection is a popular choice with customers, who say these rugs are a breeze to clean and feature a velvety smooth texture that you'll love to lounge on. The color tone has a bit of shimmer to it -- and reflects the sun in different lights for a truly stunning interior. Size options: 13 | Material: Polyester | Shapes: Motifs | Color options: 2 Best Washable Motif Rug: Norwell Moroccan Machine Woven Polypropylene Area Rug Not a fan of boring, basic rugs? Try something with a motif on it, and add a bit of color to your dreary space. This Moroccan rug from Wayfair is power-loomed from stain-resistant polypropylene -- so you won't have to worry about tracking in dirt or dropping cribs. It washes easily, and the earthy tones do a good job of masking any pet hair that might come across it (as our testing suggested). New York Post Review: "My super furry dog loves to sit at my feet or on my lap while I'm working on my laptop, so I decided it was time to try out a smaller washable rug for my desk area. This one is not just functional but super charming, too. It didn't seem to attract any of his hair, which is a huge win considering he sheds year-round. I'll warn you that it's not the softest rug out there -- the texture is a bit rougher, but it's not uncomfortable. It's on the thinner side, so I'd suggest rolling it over a rug pad for a little more support." Size options: 6 | Material: Chenille, cotton | Shapes: Medallions | Color options: 6 How We Tested Washable rugs are great investments for families with kids or pets -- as well as anyone who loves to host guests. But, they aren't all built the same. In order to find the best washable rugs in your home, we put some popular styles through several rounds of testing to ensure they were easy to care for and clean. They were washed and dried, spot-cleaned, and reviewed for comfort, as well as overall durability. * Washing and Drying Tests: We evaluated how easy each rug was to wash and dry, whether it folded easily, and how it fit into a washer or dryer. We also spot-cleaned each rug to ensure easy stain removal and vacuumed to identify whether it was easy to remove dust, grime, and pet hair. * Durability: While testing the best rugs, we observed whether there was any wear and tear from frequent traction, including loose threads or faded colors. We also considered whether the material wrinkled easily or laid smooth against the floor and if the materials were suitable for indoor or outdoor use. * Comfortability: Comfort is paramount, considering you're going to be frequently walking on your washable rug. We've lived with these rugs in our homes, taking note of the overall plushness, and thickness. softness, and quality of the material -- as well as whether they might require a rug pad. * Style and Design: Style and design are more or less subjective values, but we did consider the amount of color options and patterns that each rug came in. We also made note of the vibrancy of the tones, whether it was true to color, and how well the rug meshed with different interior spaces. Why Trust Post Wanted by the New York Post For over 200 years, the New York Post has been America's go-to source for bold news, engaging stories, in-depth reporting, and now, insightful shopping guidance. We're not just thorough reporters - we sift through mountains of information, test and compare products, and consult experts on any topics we aren't already schooled specialists in to deliver useful, realistic product recommendations based on our extensive and hands-on analysis. Here at The Post, we're known for being brutally honest - we clearly label partnership content, and whether we receive anything from affiliate links, so you always know where we stand. We routinely update content to reflect current research and expert advice, provide context (and wit) and ensure our links work. Please note that deals can expire, and all prices are subject to change.
What Anthropic's new nightmare means, in plain English
Words I did not enjoy reading this week, from a leading artificial intelligence company: "During our testing, we found that Mythos Preview is capable of identifying and then exploiting zero-day vulnerabilities in every major operating system and every major web browser when directed by a user to do so." In plain English, Anthropic says its newest AI model has found security holes in the major systems that power ... well, almost everything. Amateurs with modest coding expertise could conceivably exploit these holes to hack and crack a frightening chunk of the nation's digital infrastructure. Pulling off such a feat might be expensive in terms of computing power, but a lot of governments and non-state actors have both money and skilled hackers to spare. Maybe you suspect more AI hype, designed to goose the firm's valuation. If so, note how many major companies are taking it seriously. Instead of releasing Claude Mythos Preview to the public, Anthropic is working with a consortium of key players such as Apple, Google and Microsoft to patch these holes as soon as possible. That's a strong signal that the problem is real. So it's sobering to imagine what might have happened if a less responsible company or government had gotten to this model first -- because eventually others will get there. Anthropic may be leading the pack, but the pack is close behind. Some will see this as more reason to ban AI before it steals our passwords and our jobs. Unfortunately, that won't work, as this week's events demonstrate, because the technology is out there, and if the United States doesn't develop it, someone else will. What if the Mythos breakthrough had occurred at a Chinese firm? Any such company of sufficient importance is effectively controlled by the Chinese Communist Party, which thinks AI is very important -- so critical that the government recently barred two AI founders from leaving the country after Meta bought their start-up. If a Chinese AI developer had suddenly uncovered a wealth of security vulnerabilities, would that firm have been allowed to warn the world while helping propagate patches? Or would the exploits have been handed over to China's extensive cyberoffensive operations? https://wapo.st/4t5JVKB
Anthropic Launch Sends Fastly Down 18%, Cloudflare And Akamai Also Fall
This article first appeared on GuruFocus. Fastly (NASDAQ:FSLY), alongside Akamai (NASDAQ:AKAM) and Cloudflare (NYSE:NET), saw a sharp repricing during Friday trading after Anthropic introduced its Claude Managed Agents offering, a development that investors appear to be interpreting as a potential competitive shift in cloud and edge infrastructure. Shares moved quickly in response, with Fastly down 18%, Akamai off 13%, and Cloudflare lower by 11%, suggesting the market may be reassessing how much value could shift toward AI platforms that bundle both intelligence and execution layers. The selloff comes even as each of these platforms has been leaning into AI-driven demand. Cloudflare has been building out Workers AI and its Agents SDK, which allow developers to create agents capable of retaining context, interacting with tools, and operating autonomously. Fastly CEO Kip Compton indicated in the company's most recent earnings report that managing agents across its platform has been contributing to increased usage, while Akamai CEO F. Leighton pointed to similar dynamics within its Cloud Infrastructure Services segment. Leighton also highlighted the launch of Akamai Inference Cloud as part of efforts to support growing demand for scaling AI inference across the internet. Anthropic's Claude Managed Agents appears to extend that trajectory by offering a hosted service designed to run long-horizon agents through a simplified interface layer. According to the company, the system virtualizes key components such as sessions, orchestration loops, and execution environments, enabling each to be modified independently without disrupting the broader system. That design could potentially lower complexity for developers and enterprises, raising the possibility that AI-native platforms may increasingly overlap with functions traditionally handled by edge and cloud infrastructure providers.
SpaceX IPO Speculation Rises As SPCX Ticker Becomes Available
This article first appeared on GuruFocus. The race to get ahead of what could be one of the most closely watched IPOs in years just picked up a subtle but meaningful shift. Matt Tuttle of Tuttle Capital Management has changed the ticker on his Nasdaq-listed SPAC ETF from SPCX to SPCK, effectively freeing up SPCX a symbol that closely aligns with SpaceX at a time when speculation is building around a potential public listing that could value Elon Musk's rocket company at more than $2 trillion. SpaceX has not filed any official paperwork or confirmed a ticker, but the timing of the move could be drawing increased attention from investors watching for early signals. The development fits into a broader pattern that has been emerging across markets, where ticker symbols themselves can carry strategic value. A similar situation unfolded in 2022, when Meta Platforms (NASDAQ:META) secured the META ticker from Roundhill Investments following its rebrand, reinforcing the idea of an informal secondary market for high-profile symbols. ETF issuers have, at times, held onto intuitive tickers tied to prominent private companies, possibly anticipating future demand. In this case, Tuttle's fund launched in 2020 and holding about $7 million in assets appears to have created optionality around SPCX at a moment when investor focus on SpaceX could be intensifying. From a market perspective, ticker selection could play a role in early trading dynamics if and when SpaceX goes public. Research suggests that simple, recognizable tickers tend to attract more retail attention, which could improve liquidity and tighten spreads, at least in the near term. Still, SPCX is only one of several possible outcomes, with alternatives like X, SPAX, or MARS also circulating in prediction markets. On Polymarket, roughly $4.8 million in contracts have been traded, with Other a category that would include SPCX holding about 58% odds as of Friday, suggesting investors are still pricing in multiple scenarios.
Anthropic's New Multibillion-Dollar Deal Is a Warning Shot to Amazon and Google
Through a new multiyear agreement, CoreWeave will help Anthropic operate its Claude AI models, providing the infrastructure needed to train and deploy increasingly complex systems. The partnership will rely on advanced chips from Nvidia across U.S.-based data centers, CoreWeave CEO Michael Intrator said, according to The Wall Street Journal. Financial terms of the deal, however, were not disclosed. In the AI industry, even the most advanced model makers are becoming deeply dependent on specialized cloud providers to scale. With Anthropic onboard, CoreWeave now counts nearly all of the top AI labs as customers, including OpenAI, Meta, and Microsoft. CoreWeave's growth has been fueled by that demand. The company said Thursday it expanded a deal with Meta worth roughly $21 billion through 2032. It has also deepened its relationship with OpenAI, bringing the total value of its contracts with the company to about $22.4 billion.
Why Officials Are So Worried About Mythos, Anthropic's New AI
US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell summoned Wall Street leaders to give them an urgent warning: an artificial intelligence tool from Anthropic PBC marks the beginning of a new era of cybersecurity. The April 7 meeting in Washington was focused on Mythos, a new AI model that Anthropic says is so good at finding vulnerabilities in software and computer systems that it can only be released to a limited number of carefully-chosen parties. If tools like Mythos fall into the wrong hands, Anthropic says, it could provide attackers with a powerful new weapon to steal data or disrupt critical infrastructure. For the last several years, cybersecurity companies have promised that artificial intelligence will speed up and automate some of the work of preventing digital breaches. But hackers and cyberspies have discovered the advantages of AI too. The advent of Mythos and models like it that can exploit well-hidden flaws in popular software without human supervision points to a faster-moving, less predictable phase of the cyber arms race. What is Mythos? Claude Mythos Preview is a general purpose AI model that Anthropic says significantly outperforms prior offerings on a range of benchmarks, including for coding and reasoning. The company says it's so powerful that it has decided not to release it to the public. The company explained that some AI models have reached a level of coding capability that allows them to beat all but the most skilled humans at finding and exploiting software vulnerabilities. According to Anthropic, Mythos Preview has already found thousands of "zero-day" vulnerabilities during testing, including in every major operating system and every major web browser. "Zero days" are flaws that were previously unknown to the software's developers -- the name implying they have zero days to come up with a patch to resolve the problem. These often represent a gold mine for hackers because they offer a window of free rein inside vulnerable systems. Mythos was able to identify these with even less human intervention than past models, Anthropic said. "Mythos Preview demonstrates a leap in these cyber skills -- the vulnerabilities it has spotted have in some cases survived decades of human review and millions of automated security tests," the company said. In the hands of a ransomware gang or hostile governments, such a tool could lead to more devastating and frequent cyberattacks. Researchers say they have not been given access to independently verify Anthropic's claims about Mythos's performance. Gang Wang, an associate professor of computer science at the University of Illinois, said it's hard to assess the significance of Mythos Preview without more hands-on testing. Who will have access to it? Anthropic is calling its plan to grant access to a limited group of vetted partners Project Glasswing, after a type of butterfly with transparent wings that allow it to hide in plain sight. The participants include Amazon.com Inc., Apple Inc., Alphabet Inc.'s Google, Microsoft Corp., Nvidia Corp., Palo Alto Networks Inc., CrowdStrike Holdings Inc., Broadcom Inc., Cisco Systems Inc., JPMorganChase and the Linux Foundation, a nonprofit that supports open-source software projects. Anthropic described the project as "an urgent attempt to put these capabilities to work for defensive purposes." These organizations will use Mythos as part of their defensive security work, and Anthropic plans to share the findings of the project so others can benefit. Many companies already use so-called penetration exercises, in which they hire specialists to probe their systems for bugs so they can fix them before hackers get in. Mythos could allow companies to turbocharge that process, allowing them to find more flaws more quickly and narrow the opportunities for potential attacks. Why does Anthropic consider the release of Mythos a "watershed moment"? Anthropic described Mythos Preview as "a watershed moment for security." By their nature, zero-day vulnerabilities are difficult to find, and a small and murky industry has been built around finding them and selling them to government intelligence agencies, often for millions of dollars. According to Anthropic, the vulnerabilities Mythos Preview found were often "subtle and difficult to detect" and included a 27-year-old flaw in OpenBSD, an operating system that Anthropic says has a reputation as one of the most security-hardened in the world. Mythos was also allegedly able to turn vulnerabilities that are known but not widely patched into "exploits" that hackers could use to infiltrate computer networks. For instance, it found and chained together several flaws in the Linux kernel -- the core of the operating system and software that runs most of the world's internet servers -- to allow an attacker to take complete control of the machine. Non-experts also asked Mythos Preview to find ways to remotely take control of computers overnight and came back the next morning to a complete, working exploit, Anthropic said. Mythos is one of several new AI tools able to find zero days or build exploits. OpenAI's Codex Security and Google's "Big Sleep agent" have been developed to find vulnerabilities. OpenAI is also finalizing a product with advanced cybersecurity capabilities that it intends to release to select partners, Axios reported. Researchers at an Israeli cybersecurity startup called Buzz, meanwhile, say they have built an autonomous tool combining five AI agents that has a 98% success rate in exploiting known flaws. What safeguards are in place? The safeguards are a work in progress, according to Anthropic. "We have seen it reach unprecedented levels of reliability and alignment," Anthropic wrote, meaning it aligns with what humans want. "However, on rare occasions when it does fail or act strangely, we have seen it take actions that we find quite concerning." In one instance, a researcher urged an early version of Mythos to try to escape a secured, isolated "sandbox" computer and then find a way to send a message to that person. The tool succeeded but then continued to take "additional, more concerning actions," developing a multistep exploit to gain internet access. Anthropic said it doesn't plan to make Mythos Preview generally available, given its potential for misuse. Still, the company ultimately hopes to enable users to deploy "Mythos-class models" at scale for cybersecurity purposes and other uses. "To do so, we need to make progress in developing cybersecurity (and other) safeguards that detect and block the model's most dangerous outputs," it said. For the highest severity bugs found by Mythos, humans are involved: Specialists validate those discoveries before sending the information on to the people who maintain the code, according to Anthropic. It's a necessary but time-consuming process, but one that may eventually be eliminated as the model improves, the University of Illinois' Wang said. Does Mythos give cybersecurity defenders an advantage over hackers? Maybe, but it might take a while. Anthropic's process for disclosing flaws to the people who maintain the software or computer systems can be lengthy. So far, less than 1% of the potential vulnerabilities Mythos Preview has uncovered have been fully patched, the company said. At the same time, hackers are using AI to dramatically speed up how quickly they find and exploit vulnerabilities once they are disclosed. (Vendors are encouraged, and in some cases required, to publicly disclose vulnerabilities once they are discovered, and ideally provide a fix.) This gives cyber professionals less and less time to patch their networks. In a March 30 blog post, Palo Alto Networks Chief Executive Officer Nikesh Arora warned that the barrier for sophisticated attacks will continue to diminish over the next six months. "A single bad actor will now be able to run campaigns that required entire teams," he wrote. Yair Saban, chief executive officer of Buzz and a veteran of Israel's Unit 8200 cyber unit, said it took six engineers three weeks to build their AI-powered hacking tool. Others, including nation-state cyber spies and criminal hackers, can surely do the same, he said. Anthropic maintains that Mythos Preview and other AI tools like it will ultimately favor defenders. "In the long run, we expect that defense capabilities will dominate: that the world will emerge more secure, with software better hardened -- in large part by code written by these models," the company's Frontier Red Team said in an April 7 blog. "But the transitional period will be fraught."
Anthropic Reveals Claude Code Tool Design Philosophy Behind AI Agent Development
Anthropic has pulled back the curtain on how its engineering team designs tools for Claude Code, the company's AI-powered software development assistant. The detailed technical breakdown, published April 10, offers rare insight into the iterative process behind building effective AI agent systems. The $380 billion AI safety company's approach centers on what engineer Thariq Shihipar calls "seeing like an agent" -- essentially understanding how an AI model perceives and interacts with the tools it's given. Building Claude's question-asking capability took three attempts. The team first tried adding a question parameter to an existing tool, which confused the model when user answers conflicted with generated plans. A second attempt using modified markdown formatting proved unreliable -- Claude would "append extra sentences, drop options, or abandon the structure altogether." The winning solution: a dedicated AskUserQuestion tool that triggers a modal interface, blocking the agent's loop until users respond. The structured approach worked because, as Shihipar notes, "even the best designed tool doesn't work if Claude doesn't understand how to call it." The team's experience with task management reveals how model improvements can render existing tools obsolete. Early versions of Claude Code used a TodoWrite tool with system reminders every five turns to keep the model on track. As models improved, this became counterproductive. Claude started treating the todo list as immutable rather than adapting when circumstances changed. The solution was replacing TodoWrite with a more flexible Task tool that supports dependencies and cross-subagent communication. Perhaps the most significant shift involved how Claude finds context. The initial release used retrieval-augmented generation (RAG), pre-indexing codebases and feeding relevant snippets to Claude. While fast, this approach was fragile and meant Claude was "given this context instead of finding the context itself." Giving Claude a Grep tool changed the dynamic entirely. Combined with Agent Skills -- which allow recursive file discovery -- the model went from being unable to build its own context to performing "nested search across several layers of files to find the exact context it needed." Claude Code currently operates with roughly 20 tools, and Anthropic maintains a high bar for additions. Each new tool represents another decision point for the model to evaluate. When users needed Claude to answer questions about Claude Code itself, the team avoided adding another tool. Instead, they built a specialized subagent that searches documentation in its own context and returns only the answer, keeping the main agent's context clean. This "progressive disclosure" approach -- letting agents incrementally discover relevant information -- has become central to Anthropic's design philosophy. It echoes the company's broader focus on creating AI systems that are helpful without becoming unwieldy or unpredictable. For developers building their own agent systems, the takeaway is clear: tool design requires constant iteration as model capabilities evolve. What helps an AI today might constrain it tomorrow.
CoreWeave Signs Multi-Year Deal with Anthropic
CoreWeave, Inc. has entered into a significant multi-year agreement with Anthropic, a leader in AI research and development. This collaboration aims to support the advancement of Anthropic's Claude family of AI models. Compute resources from CoreWeave will be deployed starting later this year, enhancing the capacity for AI application globally. CoreWeave and Anthropic Partnership Overview This partnership marks CoreWeave's entry into Anthropic's robust ecosystem of infrastructure partners. It underscores the growing demand for scalable AI solutions among developers, startups, and enterprises worldwide. With this collaboration, CoreWeave now serves nine out of the top ten AI model providers, solidifying its position in the AI infrastructure landscape. Key Benefits of the Agreement * Production Scale Workloads: Anthropic will utilize CoreWeave's platform for running workloads at a production scale. * Performance and Reliability: CoreWeave offers industry-leading performance, which is essential for effective AI deployment. * Infrastructure Roll-Out: The collaboration will feature a phased rollout of infrastructure, with opportunities for future expansion. Statements from Leadership Michael Intrator, CEO and Co-founder of CoreWeave, emphasized the importance of this partnership. He stated, "AI is no longer just about infrastructure; it's about the platforms that turn models into real-world impact." Intrator highlighted the focus on deploying AI effectively in real-world scenarios through this collaboration. CoreWeave's Performance Standards CoreWeave's AI cloud is distinguished by its efficiency and high performance. It has been recognized for setting benchmark standards in the industry. Notably, it achieved an industry-leading MLPerf benchmark score for AI workloads. Additionally, CoreWeave is the only AI cloud to receive top Platinum rankings in both SemiAnalysis ClusterMAX 1.0 and 2.0, reinforcing its commitment to performance, efficiency, and reliability. About Anthropic Founded as an AI research and development company, Anthropic focuses on creating reliable and interpretable AI systems. Its flagship product is Claude, a large language model trusted by millions worldwide. To learn more, visit Anthropic's website.
Google News shows Polymarket bets in For you
Google News is beginning to surface Polymarket predictions alongside standard news stories in its personalized "For you" feed. In reported tests, Polymarket bet blocks appeared next to editorial links, and the same kind of betting-related modules were also seen on the Google News homepage. That matters because it blends two different categories of content -- news consumption and speculative prediction markets -- into a single recommendation surface. For users, that could change what they see when they open Google News: instead of headlines leading to reporting, some recommendations may point directly to betting opportunities. There's also a discoverability angle. Recommendation feeds can strongly influence what people click, and Google has a large distribution footprint across web and mobile. If betting modules keep appearing in the same places as news, Polymarket could gain additional traffic beyond its own ecosystem. From a policy and platform-risk perspective, the integration raises questions that tech newsrooms and regulators often focus on in similar situations: how platforms label sponsored or promotional content, how they separate editorial and non-editorial experiences, and how recommendation algorithms handle "predictions" that function like financial products. At the same time, Polymarket itself continues to tie its market activity to major real-world events, so being shown in a news context can feel intuitive to users who already view its predictions as an extension of event coverage. Overall, the key development is the product UI shift: prediction-market bets are no longer isolated to Polymarket's site/app, but are starting to appear within Google News' personalized browsing flow.
Bessent, Powell summon Wall Street CEOs for emergency meeting over Anthropic AI risks amid Pentagon dispute
Fox News senior national correspondent Rich Edson says the model currently 'has no plans' to be publicly available on 'Special Report.' Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell summoned Wall Street bank heads to Washington, D.C., on Tuesday for a flash meeting to warn them of cybersecurity threats posed by AI giant Anthropic, according to a Thursday night report from Bloomberg. Bessent and Powell convened the last-minute meeting at Treasury's D.C. headquarters in order to ensure the banks were ready to guard against risks from Anthropic's latest model, Claude Mythos Preview, a powerful new AI model that experts warn marks a profound shift in the technology. Each bank summoned is marked by the Fed as "structurally important" to the global financial system. The attendees included chief executives from Goldman Sachs, Citigroup, Morgan Stanley, Bank of America and Wells Fargo. Bank of America CEO Brian Moynihan was in attendance, a source with knowledge of his schedule told Fox News Digital. Spokespeople for Goldman Sachs and Wells Fargo declined to comment. Citigroup and Morgan Stanley did not immediately respond to requests for comment. PENTAGON'S AI BATTLE WILL HELP DECIDE WHO CONTROLS OUR MOST POWERFUL MILITARY TECH JPMorgan Chase CEO Jamie Dimon was also summoned but was unable to attend, Bloomberg reported, citing sources familiar. JPMorgan, notably, is a member of Anthropic's "Project Glasswing," an initiative to use Mythos as a defense against future similar models. JPMorgan did not immediately respond to a request for comment. Mythos has garnered a swell of intrigue online thanks to Anthropic's claims that the AI can autonomously identify and exploit software weaknesses. The company touted Mythos as a "frontier model" that can outperform "all but the most skilled humans at finding and exploiting software vulnerabilities." It claimed the model has already identified thousands of software flaws previously unknown to their developers, including some that were decades old inside companies widely considered to be security strongholds. "This could make cyberattacks of all kinds much more frequent and destructive, and empower adversaries of the United States and its allies," Anthropic wrote in a blog post. "Addressing these issues is therefore an important security priority for democratic states." ANTHROPIC'S DEMOCRATIC TIES UNDER FIRE AS TRUMP ADMIN SEVERS PENTAGON CONTRACTS In light of the security risks, a source close to Anthropic told Fox News Digital that the company has briefed senior U.S. government officials about Mythos, though did not specify which agencies. The increasingly relevant AI titan was once a core partner of the U.S. military, securing a $200 million contract with the Pentagon in July 2025. However, the partnership split open in February after the company drew redlines against the War Department using its technology for autonomous weapons and domestic surveillance. After issuing the company an ultimatum, Secretary of War Pete Hegseth designated Anthropic as a supply chain risk, barring federal contractors from using its products. Anthropic sought to appeal that designation, but a federal appeals court rejected their plea Wednesday. When asked to comment on the Treasury's Tuesday meeting, the Department of War referred Fox News Digital to a statement in support of the Wednesday ruling from Acting Attorney General Todd Blanche. CLICK HERE TO DOWNLOAD THE FOX NEWS APP "Today's D.C. Circuit stay allowing the government to designate Anthropic as a supply chain risk is a resounding victory for military readiness," Blanche posted on X Wednesday. "Our position has been clear from the start -- our military needs full access to Anthropic's models if its technology is integrated into our sensitive systems. Military authority and operational control belong to the Commander-in-Chief and Department of War, not a tech company." The Department of Treasury and the Federal Reserve Board did not immediately return requests for comment.
New Anthropic AI Model Deemed Unsafe For Public Release, Evens Scaring Treasury Secretary
Anthropic said the initiative formed after discovering the capabilities of Mythos Preview, stating the model "could reshape cybersecurity" but not before warning of vulnerabilities. American artificial intelligence company Anthropic is dialing back on the full release of its new AI model, Claude Mythos Preview, after revealing it is too dangerous for public usage, with U.S. Treasury Secretary Scott Bessent issuing warnings, The Hill reports. As part of Anthropic's latest initiative, Project Glasswing, Claude Mythos Preview focuses on identifying and patching security vulnerabilities in critical software. While available only to a select group of technology firms, including Microsoft, Apple, CrowdStrike, and Amazon Web Services, as well as 40 other organizations, Anthropic said the initiative was formed after discovering the capabilities of Mythos Preview, stating that the model "could reshape cybersecurity." Mythos Preview has already found several high-security vulnerabilities, including some in each major operating system and web browser, that were unknown to the software's developers prior to this, with some dating back over two decades. Before models like Mythos, these vulnerabilities would have gone undetected for years. Currently, the technology is enabling hackers and foreign adversaries to detect these vulnerabilities with ease. "Given the rate of AI progress, it will not be longer before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely," Anthropic wrote in its announcement. "The fallout -- for economics, public safety, and national security -- could be severe." Forbes outlined some findings, such as a vulnerability in OpenBSD, an operating system known as one of the most secure in the world, used to run firewalls and substantial infrastructure. It found a bug permitting anyone to crash a machine remotely just by connecting to it. After 27 years, human review missed it. It also identified a 17-year-old remote code execution vulnerability in FreeBSD, giving anyone access to a machine running from anywhere on the internet. While highlighting the dangers of these capabilities, the company also emphasizes the opportunity to find and fix software flaws and to build new software more easily with fewer security bugs. The risk spooked Bessent and Federal Reserve Chair Jerome Powell enough to call an urgent meeting with the CEOs of leading financial institutions to warn of cyber risks posed by Anthropic's latest AI model, according to Reuters. The April 7 meeting in Washington, D.C., sought to ensure that banks are aware of the risks posed by Mythos and similar models and are taking appropriate steps to defend their systems.
Anthropic's New AI Model Sparks Cybersecurity Alarm Over Hacking Risks - News Directory 3
The New York Times reported on April 7, 2026, that Anthropic describes the model as a cybersecurity reckoning. Anthropic has restricted the release of its new artificial intelligence model, Mythos, citing unprecedented cybersecurity risks. The company claims the model is too powerful to be made available to the general public because of its advanced ability to identify and exploit software vulnerabilities. According to reporting from Fortune on April 10, 2026, Anthropic is limiting access to a small group of major technology companies. These firms provide the foundational software for many other digital services, and Anthropic intends to give these defenders time to harden their systems before such capabilities are more widely available. The New York Times reported on April 7, 2026, that Anthropic describes the model as a cybersecurity reckoning. The model is capable of quickly identifying and exploiting vulnerabilities, including minor ones that would be unfeasible for humans to exploit. The potential impact of these capabilities has led to alarm regarding critical infrastructure. Reports from De Standaard indicate concerns that the model could potentially be used to hack banks and power plants, while Trends DataNews reported urgent meetings between top executives at American banks regarding the new AI model. Anthropic is not the only AI developer adopting a cautious release strategy for cybersecurity-capable models. OpenAI is reportedly developing a model internally known as Spud, which may match the capabilities of Mythos. According to a report from Axios, OpenAI also plans a phased rollout of an advanced cybersecurity-focused system to a small group of partners to provide defenders with a head start. While the companies frame these decisions as safety measures, some analysts have characterized the limited releases as a means of creating marketing hype around new models rather than purely safety-driven decisions. Regardless of the release strategy for Mythos, cybersecurity experts warn that AI-driven cyber capabilities have reached a dangerous tipping point. Fortune reports that existing, publicly available AI models can already perform sophisticated cyberattacks, sometimes within minutes. Researchers are particularly concerned about the increased scale and accessibility of these attacks. AI systems are increasingly automating tasks that previously required high levels of technical expertise, such as: This automation allows attackers who lack advanced technical skills to launch highly coordinated assaults across thousands of systems simultaneously.