News & Updates
The latest news and updates from companies in the WLTH portfolio.
SpaceX secures option to buy AI coding startup Cursor for $60B
SpaceX announced the deal on X, pre-empting a New York Times report that framed it as a completed acquisition. The structure gives SpaceX optionality: exercise the call by year end or walk away having paid $10B for shared compute access and joint model work. Cursor CEO Michael Truell called it a partnership to 'scale up Composer'. SpaceX has secured a call option to acquire AI coding startup Cursor, developed by San Francisco-based Anysphere, for $60 billion later this year, or, alternatively, to pay $10 billion for the joint AI development work the two companies are conducting together. SpaceX announced the arrangement in a post on X on Tuesday, describing "SpaceXAI" and Cursor as working together to "create the world's best coding and knowledge work AI." The post came just before the New York Times published a story citing two people who said SpaceX had agreed to purchase Cursor for $50 billion. The Times subsequently updated its story to reflect SpaceX's own framing of the deal as an option, not a completed acquisition. Cursor CEO Michael Truell confirmed the arrangement in a post on X, writing that he was "excited to partner with the SpaceX team to scale up Composer", a reference to Cursor's proprietary AI model. The option period runs through the end of 2026. Whether SpaceX exercises the $60 billion option will depend in part on how the joint model development progresses over the intervening months. No employee transfer or integration details have been disclosed. The commercial logic on both sides is clear. Cursor, a fork of Visual Studio Code with deep AI integration, developed by Anysphere, a company founded in 2022 by four MIT students: Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger, has grown at a pace that has become a benchmark for AI-era startups. It was valued at $400 million in a Series A in mid-2024, climbed to $2.5 billion by January 2025, raised $900 million at $9.9 billion in June 2025, and closed a $2.3 billion Series D in November 2025 at $29.3 billion. By February 2026 it had crossed $2 billion in annualised recurring revenue, making it the fastest B2B company to scale from zero to $2 billion in roughly three years, by widely cited metrics. More than half of the Fortune 500 now use Cursor. Co-founder and CTO Arvid Lunnemark departed in October 2025 to found Integrous Research, an AI safety lab; the three remaining founders continue to lead the company. For SpaceX, which absorbed Elon Musk's AI venture xAI in an all-stock transaction in February 2026 valuing the combined entity at $1.25 trillion, the deal addresses a visible gap. While OpenAI's Codex has reached three million weekly users and Anthropic's Claude Code has become the most-used AI coding tool among professional engineers, xAI has no comparable product. The Colossus supercomputer in Memphis, targeting one million H100-equivalent GPUs, gives SpaceX training infrastructure at scale, but without a leading application to route it through. The Cursor partnership provides that application. SpaceX had already hired two Cursor engineers, Andrew Milich and Jason Ginsberg, following an exodus of xAI co-founders. And last week, xAI began renting compute capacity to Cursor, allowing the startup to use tens of thousands of xAI chips to train its latest model, suggesting the commercial relationship predates Tuesday's announcement. The IPO context matters. SpaceX is preparing for a planned Nasdaq listing in June 2026, targeting a $1.75 trillion valuation and a raise of up to $75 billion. A $60 billion option over the world's fastest-growing AI developer tool adds narrative and commercial value to that prospectus regardless of whether the option is ultimately exercised. For Cursor, the deal provides financial certainty, either $10 billion in near-term cash for the collaboration or a $60 billion exit, without requiring an immediate sale. This is particularly notable because Cursor was simultaneously in talks as of the weekend to raise $2 billion at a valuation above $50 billion in a separate fundraising round, with Andreessen Horowitz expected to co-lead and Nvidia and Thrive Capital also participating. Whether that round proceeds alongside or instead of the SpaceX arrangement is unclear. The deal also sharpens the competitive map in AI coding tools. Cursor had previously turned down acquisition overtures from OpenAI. OpenAI's own response is to press ahead with Codex, now at three million weekly users with 40% of revenue from enterprise, and with its planned acquisition of Windsurf. Anthropic's Claude Code is the third significant player. SpaceX is now formally entering this market through Cursor's existing distribution rather than building from scratch, a faster path to relevance, at an extraordinary price.
SpaceX Pays $10B for the Right to Buy Cursor at $60B
The AI Coding Startup CEO, Michael Truell, called it a meaningful step on the startup path to build the best place to code with AI SpaceX has secured the right to purchase AI coding startup Cursor for $60 billion before the end of 2026, or walk away after paying $10 billion for nine months of joint work. The deal gives Cursor immediate access to Colossus, SpaceX's supercomputer network with compute equivalent to one million Nvidia H100 GPUs. "SpaceXAI and @cursor_ai are now working closely together to create the world's best coding and knowledge work AI," SpaceX wrote Tuesday, April 21, on X. "The combination of Cursor's leading product and distribution to expert software engineers with SpaceX's million H100 equivalent Colossus training supercomputer will allow us to build the world's most useful models. Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion or pay $10 billion for our work together." Cursor CEO Michael Truell confirmed the partnership on X, writing he's "excited to partner with the SpaceX team to scale up Composer," referring to Cursor's AI model for software development. In a blog post published the same day, Cursor explained why the company needed the deal. "We've wanted to push our training efforts much further, but we've been bottlenecked by compute," the company wrote. "With this partnership, our team will leverage xAI's Colossus infrastructure to dramatically scale up the intelligence of our models." Cursor is simultaneously raising $2 billion at a $50 billion valuation, CNBC reported Sunday, April 19, with Andreessen Horowitz co-leading the round. The deal structure means Cursor agreed to a maximum acquisition price of $60 billion while actively raising capital at $50 billion. That limits potential upside to 20 percent. The tradeoff is compute access that would cost hundreds of millions of dollars to replicate. "We released Composer less than six months ago as our first agentic coding model," Cursor wrote in its blog post. "After that, Composer 1.5 scaled reinforcement learning by over 20x. Composer 2 then added continued pretraining, reaching frontier-level performance at a fraction of the cost of other models. Each step up in compute has translated to meaningfully more capable models." The $10 billion payment structure protects SpaceX regardless of outcome. If SpaceX buys Cursor for $60 billion, the $10 billion counts toward the purchase price. If SpaceX walks away, it still receives nine months of joint development work, including access to Cursor engineers and product improvements. Cursor, founded in 2022, announced in November it had reached $1 billion in annual recurring revenue with over 300 employees. SpaceX merged with xAI in February and filed confidentially for an IPO in April. The Cursor partnership addresses a talent gap. SpaceX hired two former Cursor product engineering leads in March, Andrew Milich and Jason Ginsburg, who now report directly to Elon Musk and xAI president Michael Nicolls, according to Business Insider. At a conference in March, Musk said Grok, xAI's chatbot, "is currently behind in coding." The deal positions SpaceX to compete directly with OpenAI and Anthropic in building AI systems for software development. OpenAI was an early investor in Cursor, making the SpaceX partnership a shift in alignment. Cursor already has hundreds of thousands of developers using its tools daily, giving SpaceX immediate distribution if it exercises the $60 billion acquisition option. SpaceX can exercise the option anytime before the end of 2026. The alternative $10 billion payment gives SpaceX nine months to evaluate whether Cursor's technology and team are worth the $60 billion price tag while guaranteeing Cursor the compute access it said was preventing further model development.
SpaceX strikes major AI deal with Cursor, eyes $60 billion acquisition option The Mainstream
In a significant push into AI-driven software development, SpaceX has entered into a high-value agreement with fast-growing startup Cursor. SpaceX, the rocket and satellite company founded by Elon Musk, has announced a major deal with AI startup Cursor. The agreement gives SpaceX the option to acquire the company for $60 billion later this year or invest $10 billion in their ongoing partnership. The move is part of Musk's broader strategy to position SpaceX as a leader in AI-powered software development ahead of a potential public listing. Cursor's AI tool reshapes coding Cursor's AI assistant, launched in 2023, has changed how software is developed. It helps developers write, test, and debug code at scale. The tool has become widely used in what industry insiders call the "vibe coding" era, where AI-assisted workflows are transforming traditional coding practices. Startup growth and funding plans Founded in 2023, Cursor has rapidly emerged as one of the fastest-growing AI startups. Before the SpaceX deal, the company was in advanced talks to raise around $2 billion at a valuation above $50 billion. Andreessen Horowitz was expected to co-lead the round, with participation from NVIDIA and Thrive Capital. Musk has been actively expanding his presence in the AI space. He co-founded OpenAI, the organisation behind ChatGPT, and later launched xAI, which developed the Grok chatbot. In 2025, SpaceX moved further into AI with initiatives such as orbit-based data centres and an AI chip factory. In February, SpaceX acquired xAI in a deal valued by Musk at $1.25 trillion. Leadership comments on partnership Cursor CEO Michael Truell confirmed the partnership on social media platform X. He said he was "excited to partner with the SpaceX team to scale up Composer," referring to the company's proprietary AI model. "A meaningful step on our path to build the best place to code with AI," he added. Also read: Viksit Workforce for a Viksit Bharat Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter About us: The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.
The Consequences of SpaceX's Latest A.I. Moonshot
You're reading the DealBook newsletter. The most crucial business and policy news you need to know from Andrew Ross Sorkin and team. Get it sent to your inbox. Andrew here. SpaceX's potential acquisition of the coding start-up Cursor is raising a lot of eyebrows on Wednesday. Skeptics see it as a way for Elon Musk to shore up xAI, which SpaceX bought recently and which Musk has said needs to be rebuilt from the ground up. Some SpaceX investors grumbled about how the xAI deal diluted their stock holdings. But most will probably keep quiet if the Cursor deal happens, because they are still likely to make huge profits if SpaceX goes public this year at the $2 trillion valuation it's hoping for. SpaceX's risky coding bet Three months ago, the promise of a SpaceX I.P.O. for investors was being able to buy a piece of Elon Musk's industry-leading rocket and satellite internet business. That proposition was tested in February when the company bought xAI, Musk's money-losing artificial intelligence lab. It's happening again as SpaceX considers spending $60 billion for Cursor, a coding start-up -- another sign of Musk shifting his focus to A.I. What's happening: SpaceX wrote on social media that it had started working with Cursor to "create the world's best coding and knowledge work AI." In other words, Cursor, which has said it lacks computing resources, will get access to SpaceX's technological firepower. More important, Cursor also gave SpaceX the right to buy it for $60 billion this year. If there's no takeover, SpaceX said it would pay the start-up $10 billion -- a potentially record-setting breakup fee. The benefits for Cursor are obvious. It began 2025 as a clear leader in A.I.-assisted coding, raising $3.4 billion in funding. (Its valuation last year jumped 11-fold in 11 months, to $29 billion in November.) But in recent months, Anthropic's A.I. assistant, Claude, has rapidly seized market share in coding with new agent capabilities, and OpenAI is spending heavily to promote its Codex tool. Cursor increasingly risks falling behind, even as it's in talks to raise another round of funding at a valuation above $50 billion. It's a more mixed picture for SpaceX, however: * Grok, xAI's chatbot, isn't considered in the same tier of coding ability as Claude or Codex. Adding Cursor could change that. * But $60 billion is still a lot to spend, especially when SpaceX's debt reportedly jumped 64 percent, to $23 billion, last year. Musk claims that A.I. and space exploration go hand in hand, having talked up things like A.I. data centers in space. That said, xAI can better compete against the deeper-pocketed OpenAI and Anthropic by being attached to a publicly traded SpaceX, which could more easily tap equity and debt-financing markets. The question is whether further building out the cash-intensive xAI, which pushed the combined company to a nearly $5 billion loss last year, will make potential investors think twice about what increasingly feels like an xAI I.P.O. as much as it does a SpaceX one. HERE'S WHAT'S HAPPENING Markets whipsaw amid uncertainty about the war in the Middle East. U.S. stock futures are up on Wednesday -- but so, too, is Brent crude, the international oil benchmark -- as peace talks between the U.S. and Iran were postponed. President Trump indefinitely extended a cease-fire, but the Iranian navy claimed to have seized two vessels near the Strait of Hormuz. Lufthansa said it would cut 20,000 short-haul flights through October because of the war, and United Airlines slashed its full-year earnings guidance amid a jump in fuel costs. The Gates Foundation is reportedly planning big layoffs and an inquiry into its ties to Jeffrey Epstein. The philanthropy giant will cut up to 500 positions, or roughly 20 percent of its staff, according to The Wall Street Journal, and has opened an external review of its engagements with Epstein. The scandal could linger over the foundation as Bill Gates is set to testify before Congress in June about his relationship with the convicted sex offender. A Virginia ballot victory has bolstered Democrats' chances of retaking the House. Voters approved an aggressively gerrymandered electoral map for the state that could turn two to four congressional seats blue in time for the November midterm elections. Tuesday's results essentially nullified Republicans' efforts to protect their House majority through redistricting in states like Texas. Robinhood's venture fund buys into OpenAI OpenAI and Robinhood got into a scrap last summer, when the digital brokerage said it was offering investors access to the artificial intelligence giant's privately traded shares. They appear to have reconciled. On Wednesday, Robinhood Ventures Fund I, the firm's recently introduced publicly traded fund that invests in prominent private start-ups, is investing $75 million in OpenAI, its biggest deal yet, Niko Gallogly is first to report. The context: Last July, Robinhood offered $1.5 million worth of tokens that it said were linked to OpenAI and SpaceX. The tokens were part of the brokerage's effort to provide retail investors greater access to private companies. That irked OpenAI. "We did not partner with Robinhood, were not involved in this, and do not endorse it," OpenAI wrote on X, adding, "Please be careful." Robinhood sees a lot of retail demand for stakes in start-ups. As companies stay private longer, new investment vehicles are providing retail investors a pathway in. "The amount of growth and wealth creation that's happened in the private markets, it has just really increased dramatically," Sarah Pinto, the president of Robinhood Ventures Fund I, told DealBook. OpenAI has been courting the retail crowd. Last month, as part of its record-breaking $122 billion fund-raising round, OpenAI collected more than $3 billion from individual investors. The firm also partnered with ARK Invest to give public investors access to OpenAI's shares through three of ARK's exchange-traded funds. Adding Robinhood as an investor offers companies considering an I.P.O., including OpenAI, "a way to start building a relationship with their future shareholders," Pinto said. Broadening access to private markets can be positive for retail investors, Haim Israel, a global strategist at Bank of America, recently told DealBook. "A lot of the innovation is actually being focused in the private market, and not the public market, and investors in the retail market are missing out." But the risks are real. "You don't have ground rules to protect them," Israel said. "There always has to be a balance." On sock puppets and a potential Fed compromise Political independence. The economic impact of artificial intelligence. A headquarters renovation project that could gum up succession at the Fed. Tuesday's Senate confirmation hearing for Kevin Warsh, President Trump's nominee to run the Fed, covered plenty of ground. But it failed to give investors clarity about when the central bank might lower interest rates, or who will lead it after Jay Powell's term as chair ends next month. The futures market on Wednesday sees rates on hold until September 2027, as concerns about the inflationary impact of a prolonged war in the Middle East grow. That view is sharply at odds with Trump's. The president told CNBC before the hearing that he would be disappointed if a Warsh-led Fed didn't lower borrowing costs right away. The "sock puppet" moment: In an effort to dispel concerns about his independence, Warsh testified that the president had "never asked me to predetermine, commit, fix, decide on any interest rate decision in any of our discussions, nor would I ever agree to do so." Warsh didn't call for rate cuts on Tuesday, but he left the door open, saying that A.I. productivity gains and a smaller Fed balance sheet could justify such moves. And he largely ducked taking a position on the Justice Department's criminal investigation into Powell's handling of the renovation project or on Trump's effort to fire Lisa Cook, a Fed governor. But in response to a question from Senator John Kennedy, Republican of Louisiana, Warsh asserted that he was nobody's "human sock puppet." A big obstacle for Warsh's confirmation remains. Senator Thom Tillis, a Republican from North Carolina and a central member of the Senate Banking Committee, reiterated on Tuesday that he would block confirmation of any Fed nominee until the Justice Department's investigation ends. But Tillis supports a potential Plan B for the Fed. Some Senate Republicans have floated the creation of a special congressional committee to examine construction projects under the banking committee's purview, including the Fed's presumably. Tillis suggested he'd stop blocking the Fed nominee process under such a plan, adding, "I not only think it's a good offramp, but I also think it's good governance." Treasury Secretary Scott Bessent reportedly floated a similar idea. There's one holdup: It's not clear whether the president would be on board. "We deeply regret that this has occurred." -- Andrew Dietderich, a partner at the white-shoe law firm Sullivan & Cromwell, apologizing to a federal bankruptcy judge after his team submitted a court filing that contained artificial intelligence "hallucinations." Substack grows abroad Substack is making progress on its push into international markets, and has doubled its number of creators who charge for subscriptions in the last year, Jessica Testa of The Times is first to report for DealBook. When Substack announced it had raised $100 million last summer, its founders said they would be "doubling down" on its app, where newsletter publishers can share videos, podcast episodes and short-form posts called Notes. It was a moment of acute ambition for the company, which then had five million paid subscriptions. "I don't see anything in the laws of physics that would prevent 100 million or more paid subscriptions being on Substack one day," Chris Best, the platform's C.E.O., said at the time. The founders also teased global expansion. Of the nearly 100,000 publishers who earn money from subscriptions -- up from about 50,000 in December 2024 -- nearly 30 percent are based outside the United States. European publishers collectively earn more than $90 million annually on Substack, a company spokeswoman told DealBook. Accordingly, Substack is "nearing completion" on assembling a flock of new leaders to oversee operations in France, Germany, Italy, Spain, Denmark, Norway, the Netherlands, Sweden, Switzerland and Austria. It is also hiring, or has already made hires, in Australia, Brazil, Japan and Canada. Top-earning newsletters abroad include those written by Selvaggia Lucarelli in Italy, Zoë Yasemin and Jonas Kooyman in the Netherlands, and Jessica Troisfontaine and Lauren Bastide in France. These markets have been particularly fruitful for fashion, food and lifestyle newsletters, according to Farrah Storr, the head of international at Substack. (The top category in the U.S. remains news and politics.) But traditional media organizations, including Der Spiegel and The Guardian, have also more readily embraced the platform in Europe and the U.K. In the U.S., the largest media outlets on Substack are New York magazine and The New Yorker. Auto-translation of writing into 15 languages is also coming, first to Notes, starting on Wednesday. "It's clear that Substack is, much like in the U.S., creating a space and home for writers," including "those who dream of independence," said Storr, who previously edited the British editions of Elle and Cosmopolitan magazines. THE SPEED READ Deals * Deutsche Telekom is said to be weighing a merger with T-Mobile, its U.S. arm, potentially creating the world's biggest wireless carrier. (Bloomberg) * Sotheby's arranged a debt-financing deal with KKR through which it can borrow up to $100 million, backed by purchase fees it is owed from auction clients. (FT) Politics, policy and regulation * Letitia James, New York's attorney general, sued the crypto companies Coinbase and Gemini, accusing their prediction markets of being illegal gambling businesses. (WSJ) * "Supreme Court Appears to Back F.C.C. Fines Against Cellphone Carriers" (NYT) Best of the rest * The future of Nine West, a famous Manhattan skyscraper, seemed bleak after high-profile tenants moved out in recent years. Now it's one of the city's most desired addresses again. (WSJ) * Wagyu beef commands top prices. But what you're paying for has become less clear. (NYT) We'd like your feedback! Please email thoughts and suggestions to [email protected].
SpaceX satellites half the size of pickup trucks are falling from the sky | The Narwhal
Currently, Canada has no reporting system for space debris and no ability to limit the number of satellites launched into orbit. Existing space laws do not apply to private companies such as SpaceX and space is not covered by any environmental regulations. Billions of people watched in awe as the Artemis II mission took an astronaut crew that included Canadian Jeremy Hansen around the moon and back. It was an awe-inspiring moment for space exploration -- but not all the news from space is good for Earth. There are thousands of satellites in low orbit, which means 2,000 kilometres or less above the earth. Many were sent there by Elon Musk-owned SpaceX, which launched its first Starlink satellite in 2019 and has come to dominate the sky, representing more than two-thirds of all satellites in orbit. Wherever you are in Canada, when you look up at the increasingly bright night sky, you are seeing more satellites and fewer stars. Starlink is an internet provider used by rural farmers, northern First Nations and airplane passengers criss-crossing Canadian skies. Each of its satellites has a lifespan of roughly five years, after which they re-enter Earth's atmosphere at a rate of one or two satellites per day. At this point, they become what's known as space junk -- burning up entirely or, occasionally, scattering debris. But those occasions will become more common if SpaceX fulfills its ambitions to launch a lot more satellites in the years to come, coinciding with the explosion in data centres and artificial intelligence. That would mean more light pollution in the night sky and more space junk falling back to Earth. Samantha Lawler is a professor of astronomy with the University of Regina and goat farmer -- and she is concerned about space junk. She spoke with us from her farm in Saskatchewan (where she did not use Starlink to connect to Zoom) about why we should be concerned about the growing number of satellites over Canada -- including the potential for satellite collisions that could make low orbit unusable for everyone, a scenario called Kessler syndrome. "We're right on the edge of that already," she said, adding that someone needs to take on the engineering challenge of providing rural internet and other services with fewer satellites. "There is a limit to how many we can safely have in orbit, and I think we've crossed that limit." SpaceX didn't respond to The Narwhal's questions about the environmental or safety impacts of their plan, and the Canadian Space Agency didn't respond when asked if and when an official reporting system might be created. But Lawler had a lot more to say about the current lack of regulations protecting us from their impacts in the sky -- or here on Earth. This conversation has been edited for length and clarity. I study orbital dynamics in the Kuiper belt -- so, looking at small icy rocks in the outer solar system and measuring their orbits. I started my position at the University of Regina and moved to a farm with access to dark skies in 2019, right when the first Starlink satellites launched, so I could watch the change in my night sky that I suddenly had access to and see the change in my research data, too. Increasingly, there were more and more satellite streaks in my data. So, I had this unique perspective of seeing that wow, this was pretty bad, and it's going to get a lot worse. So, at the time, that one in 15 represented 65,000 satellites -- which, when we wrote that paper, I thought was ridiculous. Like, there's no way we'll ever get to that. But here we are at around 15,000 with no signs of slowing down. So we might get there, and now there are proposals for millions of satellites. But at the time, I think very few astronomers -- and almost no one outside the astronomy community -- had any idea how bad this was. There was a small group of astronomers that noticed, "Hey, this is very bad for astronomy. But have you thought about how many of these are going to be burning up, and how many are going to be launched, and how much danger there is in orbit?" I think that's starting to change now -- I'm glad that more people are aware of the issues, but they continue to get worse. It's so bad in every possible way. There's no way we can get to a million satellites -- there will be collisions in space and we'll be in full Kessler syndrome before we get there. But if somehow, they managed not to crash, they have five-year lifetimes. That would be one re-entry every three minutes. And those satellites would have to be bigger than Starlink satellites because of the complexity of a data centre versus an internet provider, right? In some of the articles we were writing quickly, we were estimating two tonnes per satellite, but it sounds like from various things SpaceX has released that they'll actually be much bigger than that. So everything that's in low Earth orbit, which is most of the satellites -- including all of the 10,000-plus Starlink satellites -- at the end of their life, they get burned up in Earth's atmosphere, because it's convenient. And so far, it looks like Starlink is actually doing a pretty good job of burning up. There was one piece of a Starlink satellite that was found in a farm in Saskatchewan a couple of years ago, but they seem to be doing a pretty good job. What that means, though, is that all the mass of the satellites -- the solar panels, plastic, metal, batteries -- it's all getting melted and deposited in the upper atmosphere. So, that's not a good thing. There was a period of time, about six months, where Starlink burned up 500 satellites. That's around three per day. In that time period, they exceeded the natural infall rate from meteorites by at least twice as much -- so, adding at least twice as much aluminium as what naturally comes into the atmosphere every day for six months. So what does that do? We don't actually know. There are a few preliminary studies showing this aluminum can become alumina, which can cause ozone depletion and change temperatures in the upper atmosphere, but we don't know the full effects. And because space is not legally considered an environment, all satellites launched from the U.S. are categorically excluded from any kind of environmental regulations. If they get to their steady state of having 42,000 Starlink satellites alone -- that's only one of many mega-constellations they have planned -- that's something like one satellite being burned up every hour in the atmosphere. These are satellites half the size of a Ford F-150 pickup truck. They're not small. That's a lot of metal being added to the upper atmosphere, and we don't know the full effects of it. SpaceX does all the launching -- all the other mega-constellation companies [such as One Web and Amazon's LEO] are using SpaceX to get to orbit. It has the infrastructure to do all the launches, they have a lot of U.S. government funding to do those launches, so they're doing them very, very quickly. It's very impressive engineering, it just ignores so many of the larger effects. I know in my sky, there's a line where I can always see a Starlink satellite in motion. Just always. So, people might notice that. We are also the highest-risk for satellites that aren't burning up completely, because they're right over our heads. These are all uncontrolled re-entries, so they just re-enter somewhere along their orbit, and we're under the densest part. I think that was demonstrated by the piece that was found in Saskatchewan. Actually, there are two separate things: one was a big debris fall in Ituna, Sask., which was part of the SpaceX Dragon truck. It's part of the capsule that brings astronauts up to the space station. When it doesn't burn up completely, it falls -- so that was a bunch of very large pieces discovered across many farms. I know of six pieces from that, but there are probably more that people haven't reported because there is no way to report them. There's no official reporting system. The second incident was a smaller piece from a Starlink satellite, about the size of a laptop, discovered near Hodgeville, Sask. With the Ituna debris, it was reported to the Canadian government, and there was some kind of interaction between the Canadian and U.S. governments. In Ituna, SpaceX contacted the farmers directly and came to pick up the pieces. With Hodgeville, the farmer got in touch with SpaceX, and they had him FedEx [the debris] back. So no one in the Canadian government knew about it, which is bad. The Ituna debris fall was spectacular because the pieces were so large and there were so many. But the Starlink debris is much scarier to me, because there are 10,000 of these over our heads, and if they're not burning up completely, then that's a lot of pieces that are hitting the ground. Here in Saskatchewan -- I look out my window and it's just bare fields. It's the easiest place to find the pieces. But how many pieces are we not finding? These pieces look like something that fell off a car; if you found one in the city, you wouldn't think it was space junk. Every time there's a re-entry, they just roll the dice, like, "It'll probably burn up." But we don't actually know, there's no data released on that, and the only way we find out if they aren't burning up completely is if we find pieces on the ground. I've been in touch with the Canadian Space Agency and they say they are working on a plan. But I don't know. Aaron Boley at the Outer Space Institute has set up an email address -- [email protected] -- but it's not official. We're astronomers, we're not supposed to be collecting this, but no one else is. After I heard a Starlink piece had fallen in Saskatchewan, I got in touch with the farmer by going on the Evan Bray radio show -- like, the lunchtime farmer call-in show, where I go to talk about astronomy all the time -- and asking who found it. Saskatchewan is a giant small town, so I actually got in touch with the guy by doing that. And he mentioned that his neighbour has some space junk too, and sent me a photo of this big piece of, like, corrugated metal. I was like, "Come on, that's not space junk -- it's a piece of tractor or something." But then he sent me a letter that this guy got from the Canadian government back in 1980, saying, "Thank you for sending us this piece of a Soviet rocket." So, Saskatchewan has been the debris detector for decades. Everything that goes into orbit is covered by the Outer Space Treaty and the Liability Convention, which are these Apollo-era treaties, written at a time when only the U.S. and the Soviet Union were launching stuff into orbit. They're really not written for private companies. It's just not set up for our current situation, where most of the satellites are owned by private corporations -- by one private corporation, mostly. Senator Paula Simons has launched a Senate inquiry into space junk falling on Canada, which is awesome. So there is starting to be some interest. But nothing has really happened substantively. It's hard, because Canada could say, "SpaceX, you are causing our taxpayer-funded astronomy research to suffer, so you need to pay a fine." But then SpaceX could turn around and say, "Okay, the Canadian market isn't that big, we just won't broadcast to you." A lot of Canadians are benefitting from Starlink right now -- which I don't think is a good idea, but rural internet is terrible. And then Canada would get all of the downsides and none of the upsides. SpaceX controls orbit, totally. They have two-thirds of all satellites in low orbit and if you want to go into space, you effectively have to ask them for permission. During the Artemis launch, they had all these blackout periods where there were Starlink satellites they had to avoid. By their own admission, Starlink does a collision avoidance manoeuvre every two minutes. I wrote a paper with a bunch of other people that's being reviewed, but in June, when we wrote it, it predicted that it would take five-and-a-half days for a catastrophic collision [between satellites] to happen if there were no avoidance manoeuvres. It's since dropped to three days. So if SpaceX gets hacked, or there's a bad software update, or a giant solar storm, the time we have to avoid a giant collision in orbit is getting shorter and shorter. That's a bad situation. Why does SpaceX even need 42,000 satellites to provide internet, if OneWeb is doing it with 800? They've never been asked to justify the number. We need alternatives on the ground to these internet provider mega-constellations. We need better rural internet. So something Canadians can do very easily is write to all levels of government about getting better internet to rural and remote communities, especially First Nations. I mean, no wonder everyone is using Starlink -- I live 10 kilometres from the nearest town and I can connect to power lines and phone lines and natural gas lines but I can't connect to broadband internet. That's something we can all advocate for -- because if people have good internet options based in Canada, then they don't need to rely on an American billionaire-owned company.
RBI reaches out to global regulators for risk assessment on Anthropic Claude Mythos: Report
Amid the Claude Mythos storm, the Reserve Bank of India (RBI) is reportedly engaging with international regulators, domestic lenders, and government officials to evaluate the potential cybersecurity threats posed by Anthropic's newly released AI model Mythos. According to a report from Reuters, the Reserve Bank of India (RBI) has conducted a preliminary assessment that aligns with concerns raised by regulators worldwide. Officials believe the advanced AI system could significantly accelerate the discovery and exploitation of software vulnerabilities, potentially increasing the risks to India's financial sector. RBI coordinates with global banks, regulators In the report, it is said that RBI officials have held consultations on Mythos-related risks with counterparts at the US Federal Reserve and the Bank of England, among others. The central bank may also seek direct engagement with Anthropic, the sources said. "Globally, we are discussing with other countries and other regulators on what are the developments and what safeguards need to be taken," one source told Reuters. Regulators across Asia, Europe, and the United States have already warned banks to review their defences and preparedness against potential threats from the AI model. In Japan, the financial watchdog is scheduled to meet with banks this week, while central banks in Australia and New Zealand are closely monitoring developments. NPCI seeks early access to Claude Mythos On the sidelines, India's National Payments Corporation of India (NPCI) is working alongside a small group of banks to secure early access to Mythos. The goal is to proactively identify vulnerabilities and "day-zero" cyber risks before any wider deployment from Anthropic. The NPCI currently nurtures the Unified Payments Interface (UPI) technology, which has been hailed as one of the most secure digital payment platforms available today. However, obtaining access to Mythos could be challenging, since Mythos is hosted on strictly controlled servers in the United States. Conducting tests on Mythos involving Indian customer data in foreign jurisdictions raises compliance issues, as another source raises concern. Access to Claude Mythos has so far been tightly restricted. Anthropic had initially limited the availability of Mythos to a small number of US organisations responsible for critical digital infrastructure. It is said that Anthropic plans to extend access to European banks in the near future, but nothing is concrete at the moment. RBI preparing long-term AI guidelines The RBI is also working on broader guidelines for banks entering into enterprise partnerships with advanced AI models, including Mythos and Anthropic's Claude family. These guidelines form part of a longer-term strategy to ensure a safe AI adoption in India's financial system. The discussions are still said to be in early stages, but the central bank is expected to enforce strict compliance with its 2018 data localisation rules. For those unaware, these regulations currently require payment system providers to store all end-to-end transaction data -- including user information and payment messages -- exclusively on servers located within India.
Bombay HC grants RoDTEP relief to sugar firms, flags litigation chaos
The Bombay High Court allowed RoDTEP benefits to sugar exporters and flagged inconsistent litigation by the Centre, calling for a National Litigation Policy to ensure uniformity In a ruling that could have wider implications for exporters of regulated commodities, the Bombay High Court has held that sugar shipments made with specific government permission and allocated quotas remain fully eligible for benefits under the Remission of Duties and Taxes on Exported Products (RoDTEP) scheme, even after the export policy was changed to "restricted". RoDTEP is a government scheme that refunds embedded duties and taxes on exported items to make Indian goods more competitive globally. More significantly, the court flagged "judicial chaos" caused by the Centre's inconsistent litigation strategy across High Courts and called for a clear "National Litigation Policy" to ensure uniformity on central trade and tax laws. A division bench, in a common judgment delivered on April 20, allowed a batch of writ petitions filed by major sugar exporters including Rika Global Impex, Shree Renuka Sugars, K S Commodities and Uma Exports. The court ruled that the Revenue's denial of RoDTEP benefits was arbitrary. It observed that once the government itself permitted the exports through the quota mechanism notified by the Directorate of Sugar, Department of Food and Public Distribution (DFPD), after a notification issued in May 2022, the shipments could not be disqualified from the incentive scheme merely on the ground that sugar had been placed in the "restricted" category. The dispute arose when the government shifted sugar (raw, refined and white sugar) from the "free" to "restricted" export category via the May 2022 notification to maintain domestic availability and price stability. The same notification, however, explicitly allowed exports under specific permission from the DFPD. The petitioners had exported white refined sugar within the approved quotas, fulfilled all conditions including realisation of export proceeds, and claimed RoDTEP duty credit. Customs authorities denied the benefit, citing that sugar was a "restricted" item for export and was therefore ineligible under the RoDTEP scheme guidelines. The Bombay High Court followed earlier decisions of the Gujarat High Court in Shree Renuka Sugars Ltd and Satyendra Packaging Ltd. Both those orders became final after the Supreme Court dismissed the Revenue's Special Leave Petition. The court directed the authorities to grant RoDTEP benefits to the petitioners who exported sugar with specific DFPD permissions and have not yet received them, to refund any amounts already recovered from the exporters along with 6 per cent interest per annum within four weeks, and to refrain from taking any coercive recovery action. Also Read Bombay High Court refers GST consolidated SCN issue to larger bench Consumer protection: Flat buyers' rights must not be scuttled by financierspremium No plans to impose curbs on sugar exports: Food Secretary Sanjeev Chopra Marital discord alone not basis for abetment of suicide charge: Bombay HC Bombay HC seeks Maharashtra govt's response on Muslim quota plea Advocate Abhishek A Rastogi, founder of Rastogi Chambers, who appeared for the petitioners, stated, "This judgment is a decisive reaffirmation of the principle that executive actions must be applied consistently and cannot operate to the detriment of taxpayers where the Government itself has permitted and facilitated the underlying transactions... This ruling brings much-needed certainty and restores commercial confidence for exporters operating under tightly regulated regimes." According to Anil Bhardwaj, secretary general of the Federation of Indian Micro and Small and Medium Enterprises (FISME), the Bombay High Court ruling in the sugar export matter has implications far beyond sugar, as similar issues can arise in commodities such as rice, wheat, onion, pulses, edible oils, cotton, steel products and certain minerals, where exports are often shifted from "free" to "restricted" category but continue to be allowed under quotas, licences or specific government approvals. "...For small and medium businesses, the larger concern is policy uncertainty leading to avoidable litigation and blocked export incentives. FISME believes the government should issue a uniform clarification that exports allowed under notified quotas or official permissions will remain eligible for remission benefits unless specifically excluded," Bhardwaj added. Suresh Agarwal, promoter, Rika Exports, said, "Our exports were undertaken strictly in accordance with government-approved quotas and permissions, and the subsequent denial of RoDTEP scheme benefits had created significant financial and operational strain. The High Court's decision rightly recognises that exporters cannot be penalised for policy transitions when they have acted in full compliance with the framework prescribed by the government." More From This Section Indian ports turning into logistics and industrial hubs: Shipping secy Packaging costs to dent alcobev margins by up to 200 bps, says Crisil Inflows into Indian realty jump 72% to $5.1 billion in Q1 2026: CBRE Developers acquired 3,093 acres land worth Rs 54,818 crore in 2025: JLL Capital inflows in real estate rise 72% to record $5.1 bn in Jan-Mar: CBRE
Chaos as Adrian Portelli opens his first Melbourne petrol station
Adrian Portelli has officially opened the first in his chain of petrol stations - and the event was just as chaotic as expected. The Australian billionaire launched the LMCT+ location in Preston, Victoria on Wednesday night with hundreds of his loyal followers turning up to see the spunky businessman in the flesh. Although petrol was not yet available for sale, those in attendance lined up instead for the food on offer, with Daily Mail hearing complaints of long waits and frustrated, hungry attendees. Several devotees of the luxury vehicle enthusiast showed up in their own muscle cars, with a number of drivers filling the road, revving their engines and doing burnouts outside the brand new service station. At one point, security was forced to chase off a driver sitting at a green traffic light, revving his engine and blocking the road. Portelli was mobbed by fans for several hours at the event, which culminated in one of his infamous giveaways. Adrian Portelli (pictured) has officially opened the first in his chain of petrol stations - and the event was just as chaotic as expected The Australian billionaire launched the LMCT+ location in Preston, Victoria on Wednesday night with hundreds of his loyal followers turning up to see the spunky businessman in the flesh Speaking to Daily Mail on the night, Portelli said that his aim is to serve his community by offering petrol at competitive prices as Australia faces shortages One lucky member of the public walked away from the petrol station $1 million richer thanks to Portelli's generosity. The 26-year-old announced his petrol pump venture as part of an extension to LMCT+, his subscription-based rewards club, in December 2025. Speaking to Daily Mail on the night, Portelli said that his aim is to serve his community by offering petrol at competitive prices as Australia faces shortages. 'Yeah, look, I'm gonna do what I can to relieve that,' Portelli said of the country's fuel crisis. 'I've got my customer base that have supported me over the years, so I'm willing to take a hit to the pocket to support them in these times.' Portelli indicated that he was likely to take a financial hit due to his low prices, which he intends to raise at some stage. 'This is just something to say thank you. But you know, we're gonna obviously have to at some point increase our prices, but like I said, we're gonna give wholesale fuel,' he said. The price of fuel in Australia doubled for drivers in the weeks after conflict broke out in the Middle East and Iran closed shipping passage through the Strait of Hormuz, with some service stations even running dry. Although petrol was not yet available for sale, those in attendance lined up instead for the food on offer, with Daily Mail hearing complaints of long waits and frustrated, hungry attendees Several devotees of the luxury vehicle enthusiast showed up in their own muscle cars Drivers filled the road, revving their engines and doing burnouts outside the brand new service station At one point, security was forced to chase off a driver sitting at a green traffic light, revving his engine and blocking the road Portelli was mobbed by fans for several hours at the event, which culminated in one of his infamous giveaways One lucky member of the public walked away from the petrol station $1 million richer thanks to Portelli's generosity But months before the ongoing fuel crisis began, Portelli said that his service stations would offer 'discount' fuel prices, with the first location now open in Melbourne. 'The big players in the petrol game didn't want to give you guys discounted fuel, so Adrian's gone out and bought his own to give members a whopping fuel discount,' LMCT+ said in its December announcement. 'Get ready because next year you're going to start seeing LMCT+ petrol stations roll out across Australia.' In an update last week, Portelli announced the first of the petrol stations would open within a fortnight. The project is a conversion from a Shell site on the corner of Gower Street and Plenty Road in Preston, about 10km north of Melbourne's CBD. 'The final two-week countdown!' he wrote online. 'It genuinely is surreal seeing this come to life and I got emotional seeing that LMCT+ logo hanging loud and proud!' For those struggling to find E85, which has been in short supply due to the conflict in the Middle East, Portelli added: 'There will be E85 on site in some way, shape or form.' Hundreds of Aussies applauded Portelli's decision to open a service station - with some joking his ability to find fuel in a shortage was better than that of Prime Minister Anthony Albanese. 'Bro can you replace Albo,' one wrote. Portelli indicated that he was likely to take a financial hit due to his low prices, which he intends to raise at some stage 'Yeah, look, I'm gonna do what I can to relieve that,' Portelli said of the country's fuel crisis 'I've got my customer base that have supported me over the years, so I'm willing to take a hit to the pocket to support them in these times,' he added 'This is just something to say thank you. But you know, we're gonna obviously have to at some point increase our prices, but like I said, we're gonna give wholesale fuel,' he said The 26-year-old announced his petrol pump venture as part of an extension to LMCT+, his subscription-based rewards club, in December 2025 Married At First Sight star Dave Hand (pictured) was among those who showed up My Reno Rules premiere ruthlessly mocked by viewers who have dubbed it 'The Temu Block' 'This is awesome!' another said. But others shared their doubts about LMCT+'s expansion, especially given that the brand is most famous for its lotteries. Those competitions appear to be part of the new service stations with pictures showing an on-site 'Rewards Hub'. 'Can we not bring gambling into every aspect of our lives? I'm happy for you to expand, but exposing our kids to 'a chance to win' at every opportunity is a real bad idea,' one wrote. A waitlist for LMCT+ fuel is available on the company's website but it's unclear when the first service station will officially begin sales. Portelli, from Melbourne, has an estimated worth of $1.4 billion. He made the majority of his wealth through his LMCT+ business, which exploited a loophole in gambling regulation, along with property investment and a super car collection.
RBI in talks with global regulators, banks to review risks posed by Anthropic's Mythos
Anthropic logo in an illustration | Credit: Reuters/Dado Ruvic India's central bank is in talks with global regulators, Indian lenders and government officials to understand the potential risks posed by Anthropic's new artificial intelligence model Mythos, three sources said. The Reserve Bank of India's preliminary assessment - just like that of global regulators - suggests Mythos could pose cybersecurity risks by accelerating the discovery and exploitation of software vulnerabilities, the sources, all familiar with the central bank's thinking, said. Regulators in Asia, Europe and the United States have warned banks to review defences and preparedness. In Japan, the financial watchdog will meet banks this week, while the Australian central bank said it is monitoring Mythos-related developments. RBI officials have over the past fortnight held consultations on Mythos-related risks with counterparts at the U.S. Federal Reserve and the Bank of England in particular, according to one of the sources. The RBI may seek direct engagement with Anthropic, the sources said. "Globally, we are discussing with other countries and other regulators on what are the developments and what safeguards need to be taken," one of the sources said. India's payment authority, the National Payments Corporation of India (NPCI), is trying to secure early access to Mythos alongside a small number of banks, to identify vulnerabilities and "day‑zero" cyber risks ahead of any broader rollout, this source said. However, such access may not be forthcoming as Anthropic's Mythos systems is hosted on strictly-controlled servers in the U.S. and running tests on local data in foreign jurisdictions could prove challenging, said a fourth source aware of the matter. Access to Mythos has been limited to a small number of organisations involved in maintaining key digital infrastructure in the U.S. Anthropic plans to provide Mythos access to European banks soon, Reuters reported earlier this week. Email requests for comment sent to RBI and NPCI were not immediately answered. The RBI is preparing broader guidelines for banks entering enterprise partnerships with advanced AI models, including Mythos and Anthropic's Claude family, as part of a longer‑term strategy on AI adoption, according to two of the sources. The discussions are at an early stage but the central bank will insist that all analytics based on data of Indian customers complies with RBI's domestic data localisation, the sources said. The RBI data localization rule, issued in 2018, requires all payment system providers in India to store end-to-end transaction data, including user information and payment messages, exclusively on servers located within India.
Collision Chaos: Work Vehicle Strikes Commuter Train in Washington | Business
In an unexpected early morning incident, a work vehicle collided with a stationary commuter train at the Metro Centre station in downtown Washington. This accident resulted in injuries to 11 individuals, although none were deemed life-threatening, according to reports from the Washington Metropolitan Area Transit Authority. The collision occurred shortly after midnight on the Silver Line and has since prompted the Metro to issue warnings about expected delays throughout the rail system. As a major transfer hub, the disruption affected numerous commuters relying on regular service. While the exact nature and severity of the injuries remain undisclosed, officials are focusing on determining the cause of the collision. A single track remains operational in the impacted area as investigations proceed.
SpaceX Eyes USD 60 Billion Cursor Deal Ahead of IPO, Elon Musk Pushes AI Expansion | LatestLY
SpaceX has announced a strategic partnership with the AI-powered code-generation startup Cursor, a move that includes a significant acquisition option. According to a social media announcement by SpaceX on April 21, the agreement grants the aerospace company the right to either acquire Cursor later this year for USD 60 billion or invest USD 10 billion for a collaborative partnership. The deal surfaces as SpaceX, led by Elon Musk, is reportedly preparing for an initial public offering (IPO) that could take place as early as June. The integration of Cursor is expected to bolster SpaceX's capabilities in creating high-level coding and "knowledge work" artificial intelligence, leveraging the massive computational power of the xAI "Colossus" supercomputer. SpaceX Secures Right To Acquire AI Start-Up Cursor for USD 60 Billion Ahead of IPO. The partnership is primarily driven by Cursor's need for massive computational resources to train its increasingly complex models. In a recent blog post, Cursor revealed that despite the success of its "Composer" agentic coding models, the team has been "bottlenecked by compute." By aligning with SpaceX, Cursor will gain access to the Colossus infrastructure, which SpaceX describes as equivalent to a million H100 GPUs. Cursor CEO and co-founder Michael Truell described the partnership as a "meaningful step" toward building the premier AI coding environment. The startup plans to use this newfound scale to dramatically increase the intelligence of its models, moving beyond its recent milestones in reinforcement learning and pretraining. While SpaceX's core business remains rocket launches and Starlink satellite internet, the move reflects a broader strategic shift toward artificial intelligence. Elon Musk has increasingly linked aerospace advancements with AI, recently integrating xAI into the SpaceX ecosystem and announcing plans for dedicated AI data centres and chip manufacturing. Financial analysts are closely watching the timing of this deal in relation to the anticipated SpaceX IPO. It remains unclear whether the USD 60 billion acquisition option will be exercised before or after the company goes public. The deal follows Cursor's rapid financial ascent, with the startup reaching USD 100 million in annual recurring revenue within just two years of its 2022 founding. Cursor was founded by Michael Truell, Sualeh Asif, Aman Sanger, and Arvid Lunnemark, and has already raised over USD 3 billion from private investors. Despite its rapid growth, the firm faces intense competition from established players like OpenAI and other emerging AI startups. SpaceX IPO: Elon Musk's Aerospace Giant Plans To File Prospectus As Early as This Week; Valuation Expected To Exceed USD 1.75 Trillion. The partnership with SpaceX provides Cursor with a unique competitive advantage: the specific hardware distribution and training scale required to compete with frontier-level models. SpaceX noted that the combination of Cursor's product distribution among expert software engineers and SpaceX's training power will allow the entities to build "the world's most useful models."
SpaceX's IPO could be a real problem for Tesla
That's a funny way to describe a $1.2 trillion company, but Tesla's increasingly feeling like it's playing second fiddle to SpaceX. Musk's space company, which houses his AI startup, xAI, is eyeing a historic IPO that could value it at up to $2 trillion. (He's also got a court case against Sam Altman to worry about, but that's another story entirely.) Meanwhile, the excitement around Tesla, which reports earnings this afternoon, is considerably less. The stock is down nearly 12% this year as Musk keeps pitching the world's most valuable car company as anything but a car company. From robots to robotaxis, Musk's vision for Tesla's future is very AI-focused. And it's discontinuing some car models to make room for those bets. Some progress is being made -- robotaxi service just came to Dallas -- but it's not yet bringing in meaningful revenue. Tesla's pivot is difficult to execute in the best of circumstances, let alone when your boss has an even bigger company he's preparing to take public. Musk's attention isn't the only thing Tesla is vying for. The bigger risk: investors jumping ship from Tesla to SpaceX. After all, a good chunk of Tesla investors (potentially even the majority) are more interested in backing Musk than the actual company. SpaceX represents a new way to do this, with the added benefit of being a company not in the middle of a major transformation. That doesn't mean SpaceX is a safe bet. Shooting rockets into space is a risky, expensive, and unforgiving business. And xAI's organizational structure has been in a constant state of flux, with all of Musk's cofounders having now departed. Recently, Tesla has also seen renewed interest from retail investors, according to data from Vanda Research, which is Musk's bread and butter. As crazy as this sounds, Tesla could be viewed as more of a value play (compared to SpaceX) for investors looking to invest in the Muskonomy. Still, investors love a good story. And it doesn't get much better than a space company with its eye on Mars that's on the cusp of one of the biggest IPOs.
Hackers breach Anthropic's 'too dangerous to release' Mythos AI model, report
A group of unauthorised users reportedly gained access to Anthropic's new product, which the artificial intelligence company says is too powerful to release to the public as it "poses unprecedented cybersecurity risks". Anthropic's new AI technology, Mythos, is designed for enterprise security and is being tested by a few technology and cybersecurity firms. A "private online forum" has managed to gain access to Mythos through a third-party vendor, according to Bloomberg. "We're investigating a report claiming unauthorised access to Claude Mythos Preview through one of our third-party vendor environments," an Anthropic spokesperson told Euronews Next. There is currently no evidence that Anthropic's systems are impacted, nor that the reported activity extended beyond the third-party vendor environment, the company added. Members of the unauthorised group are part of a Discord channel that seeks out information about unreleased AI models, Bloomberg reported. Citing a person employed by a third-party contractor that works for Anthropic, Bloomberg added that the group tried several strategies to gain access to the model. The outlet also reported that the unauthorised group had been regularly using Mythos once it gained access. Anthropic said it would limit the release of its new AI model to a few tech and cybersecurity firms as part of its so-called Project Glasswing. The list includes Amazon, Apple and JP Morgan Chase. Goldman Sachs, Citigroup, Bank of America and Morgan Stanley are reportedly testing the Anthropic model too, according to reports. Treasury Secretary Scott Bessent convened a meeting of senior American bankers in Washington in April to discuss the Mythos model. The meeting encouraged the banking executive to use Antropic's Mythos model to detect vulnerabilities, according to Bloomberg. Goldman Sachs, Citigroup, Bank of America and Morgan Stanley are reportedly testing the Anthropic model too, according to reports.
SpaceX secures option to buy AI coding startup Cursor for $60-billion
SpaceX said it has secured an option to either acquire code-generation startup Cursor for US$60-billion later this year, or pay US$10-billion for their new partnership, as it pushes deeper into the lucrative market for AI developer tools. Along with OpenAI and Anthropic, Cursor is one of several Silicon Valley startups that have drawn waves of developers by using artificial intelligence to automate coding, a business where AI companies have found early commercial traction. The deal could give xAI, the Grok chatbot maker that SpaceX merged with in February, a stronger foothold in the AI coding market where it has so far lagged rivals. It also provides Cursor with more computing capacity to develop AI models. "The combination of Cursor's leading product and distribution to expert software engineers with SpaceX's million H100 equivalent Colossus training supercomputer will allow us to build the world's most useful models," SpaceX said in an X post on Tuesday. Elon Musk merges SpaceX with xAI in record-setting deal valuing new company at $1.25-trillion Colossus is xAI's supercomputer cluster in Memphis, which it has touted as the largest in the world. The company has been spending billions of dollars on AI infrastructure. The announcement comes ahead of SpaceX's highly anticipated public debut in the coming months, with the company eyeing a valuation of close to US$1.75-trillion and a US$75-billion fundraise that could go down as the biggest IPO in history. Two product engineering heads at Cursor, a startup that sells AI models for coding tasks, said in March they joined SpaceX to contribute to the company's lunar projects and xAI, Musk's AI startup that is now part of SpaceX. Musk welcomed the engineers, Andrew Milich and Jason Ginsberg, saying, "Orbital space centers and mass drivers on the Moon will be incredible."
57% of Voters Believe AI Risks Outweigh the Benefits. What Does That Mean for the OpenAI and Anthropic IPOs?
Even the skeptics are benefiting from AI as the tech stocks in investors' portfolios transition into artificial intelligence stocks. OpenAI and Anthropic are both on the cusp of going public, with some investors anticipating their initial public offerings (IPOs) before the end of this year. And while those two companies are arguably the most influential software players in artificial intelligence (AI), not everyone is thrilled with AI in general. A recent NBC poll found that 57% of registered voters believe the potential risks of AI outweigh the potential benefits. Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue " While that may seem like a red flag for OpenAI and Anthropic, the companies probably don't have too much to worry about. Despite the public's concerns about AI, many investors are eager to buy AI stocks -- and most are anticipating the upcoming IPOs. Image source: Getty Images. In a recent survey from The Motley Fool, 62% of respondents said they were confident that companies investing heavily in AI would deliver strong, long-term returns. And among investors who already own AI stocks and AI exchange-traded funds (ETFs), the percentage jumps to 93%. It's not all that surprising that investors are more excited than the general public. There's a lot of uncertainty around how AI will impact many jobs, and some bosses are forcing workers to transition their workflows in ways that use more AI systems. But some investors are looking past the current difficulties and are instead focusing on the growth many tech giants have experienced lately. Here are some of the latest results for three top AI companies. Data source: Yahoo Finance and YCharts. The sale and earnings growth of these companies -- and their huge share price gains -- make it easy to see why investors may have high hopes for OpenAI and Anthropic stock, which they will soon be able to buy. Both of these privately held companies recently completed massive funding rounds that no doubt drew additional attention from investors. OpenAI raised $122 billion last month at a price that pushed its valuation up to $852 billion, and Anthropic recently raised $30 billion at a level that lifted it to a $380 billion valuation. OpenAI has 900 million weekly users, and management projects that it will have $280 billion in annual revenue by 2030. Anthropic, for its part, estimates it will have annual revenues of $150 billion by 2029. That company hasn't officially disclosed its weekly or monthly active user counts, but some estimates put total Claude users at around 30 million. There's another side to this that investors should be aware of if they aren't already: Tech investing has changed dramatically over the past few years. There isn't one tech company out there that isn't either pivoting to AI or adapting its business to be built around it in some way. And no, I'm not talking about so-called AI companies like Allbirds, which up until a couple of weeks ago was a shoe company. I mean that even if you're a tech investor (and are hesitant about AI), you're now an AI investor, like it or not. Consider that Microsoft (NASDAQ: MSFT), one of the leading software companies of the past 50 years, has invested billions of dollars into OpenAI and owns a nearly 27% stake in the company. And Alphabet (NASDAQ: GOOGL) (NASDAQ: GOOG), the tech and advertising behemoth, has one of the top AI models, Google Gemini. The company's AI already has hundreds of millions of users and will soon become the foundational model for an upcoming version of Apple's Siri, yet another tech giant shifting into AI. The point I'm trying to make is that even if the public doesn't like AI and doesn't want to invest directly in Anthropic and OpenAI after their IPOs, the tech landscape has shifted so dramatically that the 401(k)s, IRAs, and other investment accounts of the average person are now chock-full of AI companies. OpenAI and Anthropic are two of the biggest drivers of that change, and it's only going to get more difficult to differentiate between a tech stock and an AI stock. So while a large swath of the public isn't too enthusiastic about what AI will mean for them, their retirement accounts may eventually benefit from OpenAI and Anthropic's upcoming IPOs regardless. Before you buy stock in Nvidia, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now... and Nvidia wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $511,411!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,238,736!* Now, it's worth noting Stock Advisor's total average return is 986% -- a market-crushing outperformance compared to 199% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors. Chris Neiger has positions in Apple. The Motley Fool has positions in and recommends Alphabet, Apple, Broadcom, Microsoft, Nvidia, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.
Anthropic blocks entire company out of Claude without a reasonable explanation
Claude took down an entire organisation with 60+ accounts and offered no clear explanation for the decision. The CEO posted the issue on X, saying the only way to appeal the decision was by filling out a Google Form. Anthropic froze access to Claude for the entire company because of what it cited as "breaking the service's usage policy", without being any clear and only offering a Google Form to report the issue. The incident happened last week when more than 60 employees of Belo were locked out of the AI tool, Toms Hardware reported. Since they relied heavily on Claude for daily workflows, it created a major hassle for the company. They still do not know exactly which rules were broken.
Job fair disaster: Cape Town Mayor faces scrutiny following Athlone Stadium chaos
Young people wait outside Athlone Stadium for job opportunities. Cape Town Mayor Geordin Hill-Lewis is facing mounting pressure to account for the chaos at Athlone Stadium, where an estimated 25 000 job seekers descended on the City's "A Day of 1 000 Opportunities" event, triggering a stampede that left several people injured. According to reports, the initiative, aimed at linking unemployed youth to opportunities in the contact centre industry, drew crowds far exceeding the venue's capacity. Many hopefuls had queued from as early as 4am, only to be turned away once gates were shut. The event was hosted by CapeBPO, the City's special purpose vehicle (SPV), as part of efforts to create a direct pathway into the growing sector. ANC councillor Judy-Ann Stevens said the mayor must take responsibility for what she described as a serious failure in planning and execution. "You cannot invite thousands of desperate unemployed youth to a jobs initiative without proper crowd control, safety measures and clear communication," she said. She added that many attendees spent money they could not afford, endured long queues, and were ultimately met with confusion, overcrowding and injuries and, in some cases, arrests. "We are calling on the mayor to take full responsibility. This... reflects broader issues in how the City engages with unemployed youth," she said. Hill-Lewis referred questions to Mayco member for economic growth, James Vos. Responses were later provided by the City's economic development directorate. "As per the City's update last week, immediate steps were taken by stakeholders to assist with the high volumes of applicants, with thousands of attendees receiving assistance at the venue," the City said. "Medical personnel were on the scene to assist those who sustained injuries due to crowd jostling. "The City thanks all teams, including emergency services, Law Enforcement, SAPS and all personnel who assisted on the day to cope with the high turnout, which far exceeded the previous four years of this annual event, and the City regrets that not everyone could be attended to on the day." Stevens also questioned the effectiveness of such job-seeker programmes. "A similar initiative was held last year and many young people registered and are still waiting for responses. So the key question is: are they actually creating jobs or false hope? "Job creation initiatives must be dignified, properly planned and must lead to real outcomes -- no chaos and disappointment. We are calling for a full investigation, support for those who were injured and assurance from the City that this will never happen again," she said. The City said more than 100 000 jobs have been created by the BPO industry. "The BPO sector continues to present significant opportunities for Cape Town residents, employing over 100 000 people, especially in lower-income communities. "Industry growth continues to be driven by CapeBPO, the City's special purpose vehicle for the call centre industry."
How to Stay Calm When Your Classroom Feels Like Chaos
You're halfway through explaining fractions when three students start arguing, someone's throwing pencils and the noise level rivals a food court at lunch. Your heart rate spikes. Your jaw tightens. And you're wondering how you're supposed to teach anything when you can barely think straight. The data paints a troubling picture. According to the 2023 Monash University Teacher Survey, nearly 90% of teachers reported feeling burnt out or stressed by their workload. Research from the University of Melbourne and the ANU confirms that teachers experience psychological distress at rates significantly higher than the general population, scoring three times higher for depression and nearly four times higher for stress compared to national norms. The classroom chaos isn't just disrupting your lessons. It's affecting your mental health. But staying calm in a chaotic classroom isn't about having supernatural patience or pretending everything's fine when it's not. It's about building systems that prevent the chaos before it starts and having strategies ready when things go sideways anyway. Build Your Calm Before the Storm Hits The best time to manage classroom chaos is before students walk through the door. First, start with clear, visible routines that students can follow without asking. When students know exactly what happens when they enter the room, where to put their bags and what to do in the first five minutes, you eliminate dozens of small decisions that create noise and confusion. Next, post your daily schedule where everyone can see it. Use the same structure each day. Students thrive on predictability and you'll feel calmer when you're not constantly fielding questions about what happens next. Additionally, teach your expectations like you teach math. Don't assume students know how to line up quietly, transition between activities or work in groups. Model it. Practice it. Reinforce it. In fact, research from Victoria's evidence-based framework shows that behaviour can be taught through explicit teaching and positive reinforcement. Disciplinary measures alone don't teach students what to do. Ultimately, when you invest time upfront teaching these behaviours, you save yourself hours of stress later. Use Preventive Strategies That Actually Work You can't control every variable in your classroom, but you can stack the deck in your favour. First, position yourself strategically. Move around the room constantly. Stand near students who need proximity to stay focused. Your physical presence prevents problems before they escalate. Similarly, use non-verbal signals to redirect behaviour without interrupting your lesson. A look, a gesture or a tap on the desk. These quiet corrections keep you calm because you're not raising your voice or stopping to address every small issue. Furthermore, build in brain breaks. When you notice energy building, pause for 60 seconds. Have students stand, stretch or do a quick movement activity. This releases tension for them and gives you a moment to reset. According to PISA data analysed by ACER, Australia ranks below the OECD average for classroom disciplinary climate, with 43% of Australian students reporting noise and disorder in their science classrooms. Australia has one of the noisiest classroom environments in the developed world. Students learning in favourable classroom environments scored between 40 and 81 points higher in mathematics than those in less favourable environments. This represents roughly one to two years of schooling. Clearly, your calm matters for their learning. Respond to Disruptions Without Losing Your Cool Even with the best prevention, disruptions happen. How you respond determines whether the situation escalates or de-escalates. First, lower your voice instead of raising it. When the room gets loud, speak more quietly. Students have to quiet down to hear you. This counterintuitive approach keeps your stress hormones from spiking and models the calm behaviour you want to see. Second, use the pause. When a student acts out, take three seconds before responding. This tiny gap prevents you from reacting emotionally and gives you time to choose a measured response. Third, address the behaviour, not the student. Say "We need quiet voices during independent work" instead of "You're being too loud." This removes the personal confrontation that triggers defensiveness in students and frustration in you. Keep a simple phrase ready for tense moments. Something like "Let's reset" or "Back to learning mode." This gives you a neutral script when your brain is too stressed to improvise. Finally, deal with major disruptions privately. Don't have a confrontation in front of 25 witnesses. A quiet "I need to speak with you for a moment" followed by a brief hallway conversation prevents the power struggle that drains your energy and disrupts everyone's learning. Protect Your Capacity to Stay Calm You can't pour from an empty cup and you can't stay calm when you're running on fumes. More than two-thirds of Australian teachers describe their workload as largely or completely unmanageable. Administrative duties and compliance requirements take time away from lesson planning and student engagement. This matters because your ability to stay calm in the classroom depends on your overall stress levels outside of it. First, set boundaries around your time. Leave school at a reasonable hour most days. Your evening prep work can't come at the cost of sleep, exercise or basic self-care. Second, simplify your systems. Use the same lesson plan template every week. Create one grading rubric you can adapt for multiple assignments. Batch similar tasks together. Every decision you eliminate during your workday preserves mental energy for the moments when you need it most. Finally, connect with other teachers. Share strategies. Vent when you need to. Celebrate small wins. Teaching can feel isolating, but you're not alone in this struggle. Remember What You're Actually Managing The goal isn't a silent classroom where students sit perfectly still. Instead, the goal is a learning environment where you can teach and students can focus without constant disruption. Some noise is normal. Your job is to keep the chaos from tipping into disorder that prevents learning. When you use preventative strategies, respond calmly to disruptions and protect your own well-being, you create the conditions for both you and your students to succeed. Start with one strategy from this article. Master it. Then add another. However, your classroom won't transform overnight, but your stress levels will start to shift as you build these habits into your daily practice. You've got this
Unauthorised users accessed Claude Mythos, Anthropic investigates
Anthropic's decision to tightly restrict access to its newly announced Mythos AI model has come under scrutiny after reports that unauthorised users were able to interact with the system almost immediately after its launch. According to a Bloomberg News report, a small group operating within a private online forum accessed the model on the same day it was introduced for limited testing. The model, which is not publicly available due to concerns around its potential misuse in cybersecurity, was meant to be deployed only to a select group of approved organisations. Access through third-party environment The report indicated that the breach was linked to a third-party contractor environment rather than Anthropic's core systems. Users in a private Discord group are said to have combined multiple methods to gain entry, including the use of tools typically associated with cybersecurity research. They reportedly identified the likely hosting location of the model by analysing patterns from earlier Anthropic systems, along with clues drawn from publicly available information and past incidents involving related platforms. The group is also known to track unreleased AI models and scan platforms like GitHub for indicators. The users have continued to access and interact with the Mythos model since gaining entry, the report said. However, their activity has not been focused on cybersecurity use cases, despite the model being designed specifically for defensive applications. Anthropic's response An Anthropic spokesperson said the company is examining the situation. "We're investigating a report claiming unauthorized access to Claude Mythos Preview through one of our third-party vendor environments," the spokesperson said. The company added that it has not found any evidence to suggest the access extended beyond the third-party environment or affected its internal systems. Why Mythos is tightly controlled Anthropic introduced Mythos on April 7 under "Project Glasswing," a controlled programme aimed at allowing select organisations to test the model for defensive cybersecurity purposes. During testing, the model reportedly identified thousands of critical vulnerabilities, including zero-day flaws that would typically take human researchers months to uncover. Experts told Business Insider that the system can significantly reduce the time required to develop exploits, highlighting both its capabilities and associated risks. The company has limited access due to concerns that such capabilities could destabilise the cybersecurity landscape if misused. It has opted for a restricted rollout instead of a public release. Limited access, growing interest A small number of companies, including Amazon, Apple and Cisco Systems, have been allowed to test the model. It is also being made available to select organisations through Amazon's Bedrock platform. The report added that financial institutions and government agencies are seeking early access to better assess the risks and defensive applications of the technology.
Kraken commences K3 USV production with Rheinmetall, partners with Anduril
An image of Kraken's K3 Scout USV accompanying Rheinmetall-Kraken's announcement of series production of the K3 on 20 April 2026. The image was likely taken during a demonstration in Hamburg in March 2026. (Rheinmetall Naval Systems) Rheinmetall Kraken GmbH has commenced series production of the K3 Scout unmanned surface vehicle (USV) at its Blohm+Voss site in Hamburg, Germany, the company said in a press release on 20 April 2026. In the press release, Tim Wagner, CEO of Rheinmetall's Naval Systems division said that "production of the Kraken K3 Scout is initially planned for around 200 units per year. Depending on order volume, we can ramp up production to up to 1,000 units annually." K3 production is intended to address "the increased global demand for commercially available unmanned platforms of varying sizes", the press release said. K3 is intended as a low-cost and low-signature USV featuring composite construction and diesel propulsion. The USV features a modular payload bay and can be configured according to customer requirements. The K3 Scouts in production in Hamburg are 8.5 m in length and capable of reaching speeds of 55 kt, indicating the K3 Scout Medium variant of the USV. Other versions of the USV include the 11 m K3 Scout Heavy and the 18.6 m K3 Scout Max. K3 can be used for civilian or military tasks including intelligence, surveillance, and reconnaissance (ISR), logistics, search-and-rescue (SAR), survey tasks, anti-surface warfare (ASuW), counter-unmanned aerial system (C-UAS) missions, one-way attacks, and other kinetic missions. The Rheinmetall Kraken joint venture (JV) was originally announced between German shipyard Naval Vessels Lürssen (NVL) and British company Kraken Technology Group on 22 August 2025. In September 2025, it was announced that Rheinmetall was to acquire NVL with the transition concluded on 1 March 2026.
