News & Updates
The latest news and updates from companies in the WLTH portfolio.
Palantir CEO criticizes OpenAI, Anthropic over AI token value concerns
https://scoop.upworthy.com/billionaire-reveals-hes-supporting-landlord-to-sheltered-him-and-mom Palantir CEO Alex Karp has publicly criticized OpenAI's Sam Altman and Anthropic's Dario Amodei, accusing them of charging Fortune 500 companies for AI tokens that allegedly offer no real value. In a recent interview, Karp suggested that these tech leaders are exploiting enterprises by charging for services that fail to deliver tangible business outcomes while potentially misusing proprietary data. This criticism comes amid growing dissatisfaction among enterprises with the high costs and low returns from token-based AI models. Karp's comments highlight ongoing tensions in the AI sector, where companies are increasingly skeptical of the current pricing models offered by leading AI labs. Key Takeaways * Palantir CEO Alex Karp's remarks appear to suggest skepticism over the value and honesty of token-based AI services provided by OpenAI and Anthropic. * Market pricing indicates a potential impact on Anthropic's valuation, with the likelihood of hitting high targets by December 31 showing varied confidence. * The broader AI market is experiencing a shift in sentiment as enterprises reconsider the return on investment from token-based AI models. What to Watch Markets may closely monitor how Anthropic and OpenAI respond to Karp's allegations, which could influence future enterprise contracts and partnerships. Any official statements or strategic shifts from these companies could further affect Anthropic's market valuation trajectory. Additionally, watch for any changes in investment activity or partnership announcements, particularly from major investors like Amazon and Google, which could provide further insight into market confidence. Get live prediction-market analysis, powered by Vera. Sign up for Vera.
Anthropic rumored to unveil AI model surpassing GPT-5.6 SOL next week
A speculative post on social media suggests that Anthropic could release a new, more affordable AI model next week that surpasses OpenAI's GPT-5.6 Sol in intelligence and cost. The claim, made by user @cryptopunk7213, remains unverified as there has been no official confirmation from Anthropic. Currently, Anthropic's leading model is Claude Fable 5, available globally since July 1, following the easing of export controls. This speculation comes amid intense competition in the AI sector, with OpenAI's GPT-5.6 Sol leading recent benchmarks and priced more competitively than its rivals. Key Takeaways * The social media claim suggests Anthropic might introduce a model that outperforms GPT-5.6 Sol, affecting AI market dynamics. * Market pricing indicates a potential increase in confidence in Anthropic's ability to secure a top position in the AI landscape. * The claim lacks official corroboration, emphasizing the speculative nature of the information. What to Watch Monitor Anthropic for any official announcements or developments regarding a new AI model release. Any confirmation from Anthropic could significantly impact the current market odds, which appear supportive of Anthropic securing a top-three position by the end of July 2026. Additionally, observe reactions from competitors like OpenAI, which may respond with strategic adjustments or announcements to maintain their competitive edge. Get live prediction-market analysis, powered by Vera. Sign up for Vera.
Polymarket predicts 57% chance of no rate change in September
Prediction market traders are betting the Fed holds steady despite 4.2% inflation, but a 37% chance of a hike keeps things interesting The crowd has spoken, and the crowd thinks the Federal Reserve is going to sit on its hands. Polymarket, the blockchain-based prediction market, currently prices a 56% probability that the Fed will leave interest rates unchanged at the September FOMC meeting. Shares representing a "no change" outcome are trading at 57 cents each, with the market generating over $2 million in trading volume since it launched on May 13. That's not trivial engagement for a single policy question months in advance. The rate hike that might not happen While the majority bet is on the Fed doing nothing, a 25 basis point increase sits at 37% probability -- roughly one in three traders thinks the Fed will actually raise rates further. The May Consumer Price Index came in at 4.2% year-over-year, well above the Fed's 2% target. The labor market, meanwhile, continues to show resilience. Broader 2026 Fed policy markets on Polymarket tell a similar story. The probability of any rate hike occurring this year sits in the 51-53% range, making it essentially a coin flip on whether the Fed moves at all. Polymarket's track record Polymarket priced a 25 basis point cut in September 2025 at 91% confidence, and the Fed followed through. The market resolves based on official FOMC statements and Federal Reserve data releases, rounding to the nearest 25 basis points. The platform operates on the Polygon blockchain and accepts USDC or pUSD as collateral. Trading volume of over $2 million for a single FOMC meeting outcome signals that this isn't just retail speculation, with the platform serving as a complement to the CME FedWatch tool that has long dominated rate probability forecasting. What this means for crypto investors Rate decisions are one of the single biggest drivers of risk asset prices, and crypto remains firmly in the risk asset category. A rate hold in September would likely be interpreted as neutral-to-positive for crypto, as no tightening means no additional pressure on liquidity. The 37% chance of a rate hike is the scenario crypto investors should game plan for. Bitcoin and Ethereum have historically shown sensitivity to unexpected hawkish pivots. If subsequent CPI prints come in hotter than 4.2%, that 37% hike probability could climb, and crypto markets would likely start pricing in the pain before the Fed even acts.
Kraken leads with $400M in spot liquidity across MiCA exchanges
DefiLlama's new compliance dashboard shows Kraken pulling ahead of Coinbase and rivals as Europe's crypto regulation fully kicks in Kraken is sitting on roughly $400 million in spot liquidity across MiCA-licensed exchanges, making it the clear frontrunner in Europe's newly regulated crypto marketplace. The exchange also holds about $207 million in perpetual liquidity, putting meaningful distance between itself and every other compliant competitor on the continent. The numbers come from DefiLlama's freshly launched MiCA compliance dashboard, which went live on July 1, 2026. That date coincides with the full enforcement of the Markets in Crypto-Assets Regulation for crypto asset service providers across the European Union. The liquidity leaderboard takes shape The exchange's $399.71 million in spot liquidity dwarfs its nearest competitor. Coinbase, the second-place finisher, reports approximately $305 million in spot liquidity and $167 million in perpetuals. That's a roughly $95 million gap in spot alone. From there, the drop-off gets steep. Crypto.com trails with around $131 million in spot liquidity. Bitstamp, one of Europe's legacy exchanges, comes in at roughly $55 million. And OKX sits near the bottom with about $12 million in spot liquidity and lower or no perpetual liquidity to speak of. Kraken's platform currently supports trading across 1,704 markets, covering both spot and perpetual products. Why MiCA changes the game The Markets in Crypto-Assets Regulation represents the EU's attempt to build a single, unified licensing framework for crypto service providers. One license, one set of rules, access to all 27 member states. Kraken moved early. The exchange secured its MiCA authorization from the Central Bank of Ireland back in June 2025, a full year before the regulation's enforcement deadline for crypto asset service providers. That head start gave Kraken time to build out its European operations, including spot, futures, and derivatives offerings, while competitors were still working through the licensing process. DefiLlama's MiCA compliance dashboard lets users compare exchanges on liquidity, compliance status, and transaction fees, all in one place. What this means for investors Coinbase, sitting in second place with $305 million in spot liquidity, remains a formidable competitor. But the $95 million gap to Kraken is significant enough to influence where large orders get routed. Smaller MiCA-licensed platforms like OKX, with just $12 million in spot liquidity, face a difficult question: can they grow fast enough to remain viable, or will they become acquisition targets for larger players looking to expand their European footprint?
Grok gets creative tools as Musk pushes xAI deeper into image and video generation
The integration of Imagine into Grok's agentic mode signals xAI's aggressive expansion into autonomous content creation, with potential ripple effects across digital asset markets. Elon Musk just gave his AI chatbot a paintbrush. And a camera. And, if he's to be believed, eventually a director's chair. On July 8, Musk announced that Grok, the conversational AI built by his company xAI, will integrate the Imagine tool for autonomous image and video generation. The key detail: Grok will be able to call Imagine dynamically in "agentic mode," meaning the AI decides on its own when to generate visual content rather than waiting for explicit user commands. What Grok's agentic mode actually means The Imagine tool itself has been evolving rapidly. Grok Imagine Video 1.5 launched around June 16-17, bringing what Musk described as significant improvements in both quality and speed. Some updates have enabled image-to-video generation in under 15 seconds. Musk predicted that Grok could produce full movies by the end of 2026. The AI content creation arms race The competitive landscape in AI-generated video has intensified dramatically over the past year, with OpenAI's Sora, Google's Veo, and a constellation of smaller startups all racing to build tools that can produce broadcast-quality video from text prompts. What makes the Grok-Imagine integration notable isn't just the technology itself. It's the distribution. Grok lives inside X (formerly Twitter). If Imagine's video generation reaches even a fraction of that audience through agentic mode, xAI instantly becomes one of the largest deployments of AI video generation in the world. Where crypto and digital assets enter the picture Musk's announcement contained zero references to cryptocurrency, blockchain, tokens, or any digital asset infrastructure. Not a single one. xAI isn't launching a token. Imagine isn't minting NFTs. There's no on-chain component to any of this.
Polymarket races to regain trust with US market re-entry after $112M acquisition
The prediction market giant is betting that rock-bottom fees and a CFTC-approved derivatives exchange can make Americans forget it once kicked them off the platform Polymarket, the crypto-native prediction market that became a cultural phenomenon during the 2024 US election cycle, is making its way back to American users. The platform acquired regulated derivatives exchange QCX for $112 million, giving it the legal scaffolding to offer event contracts stateside. Polymarket didn't leave the US market on its own terms. Back in 2022, the platform was forced to block American users after running into regulatory headwinds. Now it's attempting a comeback, armed with federal approval and a fee structure aggressive enough to make both offshore venues and traditional sportsbooks uncomfortable. The regulatory path back The company secured an amended CFTC order in November that allows it to beta test its US exchange with live trades. The QCX acquisition is the centerpiece of this strategy. By purchasing a regulated derivatives venue for $112 million, Polymarket essentially bought itself a compliance moat rather than trying to convince regulators that prediction markets deserve a novel framework. The US-specific app will initially focus on sports contracts. Politics and crypto markets are planned for later. A fee structure designed to hurt competitors Polymarket's US product features 10 basis point taker fees with zero maker fees. Standard sportsbook vigorish typically runs anywhere from 4% to 10% on most bets. A trader placing a $1,000 position on Polymarket pays $1 in fees. The same economic exposure through a sportsbook would cost somewhere between $40 and $100 in embedded margin. The zero maker fee component means liquidity providers can operate for free, bootstrapping deep order books. Why the trust problem is real American users who were active on the platform before 2022 remember being shown the door. Some found workarounds, using VPNs and non-US accounts, which created its own set of problems when questions about market integrity surfaced during the 2024 election. The fact that US participation was technically prohibited while the platform was being used to forecast US elections created an awkward dynamic that regulators noticed. Now Polymarket has to convince American retail users the platform won't pull the rug again if regulatory winds shift, and institutional participants need to see a compliance infrastructure robust enough to satisfy their own legal teams. A $112 million price tag for a regulated venue signals commitment that's hard to fake. What this means for the broader prediction market landscape Kalshi, the other major US prediction market, has been operating with CFTC approval since 2020 and has fought its own regulatory battles to expand into election contracts. Polymarket's entry as a direct competitor with dramatically lower fees could force Kalshi to reconsider its own pricing. Analysts watching this space see Polymarket's return as a potential inflection point for on-chain prediction markets more broadly. If a CFTC-approved platform with institutional-grade compliance can operate at 10 basis point fees, it validates the entire category as a legitimate financial product rather than an unregulated gray area.
Anthropic expands in New York City with massive 466,000 sq ft office lease
The Claude AI maker is taking over an entire 16-story building in Lower Manhattan, signaling aggressive growth that intersects with crypto-adjacent infrastructure deals Anthropic, the AI company behind the Claude model, is leasing an entire 16-story building at 330 Hudson Street in Lower Manhattan. The deal covers roughly 466,000 square feet of office space, a staggering 30x increase from the company's current New York footprint. For context, Anthropic currently occupies about 15,500 square feet at 155 Avenue of the Americas, a space it leased in 2024. The current lease at 155 Avenue of the Americas has a potential expiration approaching in 2026, making the timing of this deal practical as much as aspirational. From startup footprint to tech giant ambitions Anthropic had been shopping for between 250,000 and 450,000 square feet of Manhattan office space since as early as January 2026. Landing at the top end of that range tells you something about how quickly the company's ambitions scaled during the search process. The company plans to double its New York workforce as part of the move, as confirmed in a July 7, 2026 report by the New York Times. The TeraWulf connection and why crypto investors should pay attention Anthropic recently signed a separate $19 billion, 20-year lease agreement with TeraWulf for AI data center infrastructure in Kentucky. TeraWulf started life as a Bitcoin mining company. It built out substantial power infrastructure and data center capacity to mine cryptocurrency, then increasingly pivoted toward hosting AI workloads as the economics shifted. The company essentially realized that the same cheap power and cooling infrastructure that makes Bitcoin mining profitable also makes it ideal for running the massive GPU clusters that AI companies need. The risk is concentration. A $19 billion, 20-year commitment to a single tenant means TeraWulf's fortunes are now deeply tied to Anthropic's success. Investors should watch whether TeraWulf maintains a balanced portfolio of AI and crypto mining clients or becomes overly dependent on one relationship.
SpaceX's lead underwriters face $1T valuation gap as quiet period ends
Goldman Sachs and Morgan Stanley released wildly divergent revenue forecasts for the newly public space giant, and the difference comes down to one word: AI When two of Wall Street's most powerful banks can't agree on what a company is worth, and the gap between their estimates stretches to roughly $1 trillion, investors should probably pay attention. SpaceX's post-IPO quiet period ended in early July, unleashing a flood of analyst reports from the underwriters who shepherded the largest public offering in history. Goldman Sachs and Morgan Stanley, the two lead underwriters, published their inaugural coverage notes within days of each other. The numbers that don't add up Goldman Sachs projects SpaceX will generate $474 billion in total revenue by 2030. Morgan Stanley pegs that figure at $330 billion. That's a $144 billion disagreement on a four-year outlook. The core of the disagreement sits squarely on AI. Goldman attributes $322 billion of its 2030 revenue estimate to AI operations, while Morgan Stanley sees that segment contributing $190 billion. A $132 billion gap in a single revenue line item. Zoom out to 2040 and the divergence gets genuinely absurd. Morgan Stanley forecasts $3.4 trillion in revenue and over $2.7 trillion in adjusted EBITDA by that year. If Goldman's more aggressive growth assumptions hold through the decade, the implied valuation gap between the two banks' models balloons to around $1 trillion. The IPO that broke records SpaceX priced its IPO at $135 per share on June 11, 2026, raising $75 billion in one of the most anticipated public offerings ever. The greenshoe option pushed the total raise to $85.7 billion. Only about 4% of the company was sold to the public. When trading began on June 12, shares surged enough to push SpaceX's market capitalization to approximately $2.1 trillion. The implied equity valuation at offering was about $1.77 trillion, meaning the market added roughly $330 billion in perceived value on day one alone. The underwriting fees tell their own story. At under 0.75% of the total raise, SpaceX negotiated a fee structure well below the typical 3-7% charged on large IPOs. Even so, the sheer size of the deal meant the total fee pool landed somewhere between $500 million and $650 million. Goldman Sachs and Morgan Stanley each captured approximately 20% of that pool, meaning each bank walked away with north of $100 million for their efforts. Why the AI bet matters for everyone The two banks are effectively placing opposite-end bets on how quickly AI capabilities can be monetized at scale through satellite infrastructure and space-based computing. Goldman's model assumes AI operations become the dominant revenue driver within four years, essentially dwarfing the launch and satellite connectivity businesses that made SpaceX famous. Morgan Stanley's model treats AI as a significant but not overwhelming contributor, keeping more weight on legacy revenue streams. Investors watching SpaceX should focus on the quarterly AI revenue disclosures that will begin arriving later this year. The first few earnings reports will start revealing which bank's crystal ball is less foggy. If AI revenue tracks closer to Goldman's projections, the stock likely has room to run. If Morgan Stanley's estimates prove more accurate, the current $2.1 trillion market cap could face pressure.
Anthropic to invest $15B in Australian data centers by mid-2027
https://itbrief.com.au/story/anthropic-plans-australian-office-in-global-ai-push Anthropic has announced plans to invest up to $15 billion in developing data centers across Australia by mid-2027. The move is part of a strategic expansion to secure 1.4 gigawatts of data center capacity, primarily through leasing from local developers or joint ventures. This development follows an earlier memorandum of understanding with the Australian government and aligns with Anthropic's broader strategy to establish Western-aligned infrastructure outside the United States. The company, known for its AI model Claude, aims to activate at least 1 gigawatt of capacity by the end of 2027, supporting its future AI model training and deployment efforts. Key Takeaways * The announcement appears consistent with increasing valuation expectations for Anthropic, as it reflects significant growth potential. * Markets suggest that the $15 billion investment could be a key indicator of Anthropic's commitment to expanding its infrastructure footprint outside the U.S. * Current pricing in related prediction markets is supportive of scenarios where Anthropic's valuation increases significantly by the end of the year. What to Watch Observers will be closely monitoring Anthropic's progress in securing the planned data center capacity in Australia, particularly through partnerships with local developers. Developments that indicate successful leasing or joint ventures could be consistent with a YES outcome in valuation increase scenarios. Additionally, any further announcements regarding Anthropic's strategic initiatives or partnerships with major tech firms like Amazon and Google could influence market expectations. Get prediction market intelligence as a structured API feed. Early access waitlist.
Meta Platforms restricts engineers from using Anthropic's Claude and OpenAI's Codex
The tech giant is locking down rival AI tools to prevent 'model distillation' risks, joining Microsoft in a broader industry pullback from third-party coding assistants Meta has drawn a hard line in the AI arms race. The company's Applied AI division now explicitly prohibits engineers from using Anthropic's Claude Code and OpenAI's Codex, two of the most popular AI-powered coding tools in the industry. The restrictions, documented internally around June 29, 2026, are designed to prevent a specific and somewhat ironic problem: rival AI models accidentally teaching Meta's own models their tricks. The distillation problem The concern is straightforward. When Meta engineers use Claude Code or Codex to write code, generate data, or run evaluations, those outputs carry the DNA of Anthropic's and OpenAI's proprietary models. If that output then feeds into Meta's training pipelines, evaluation benchmarks, or post-training datasets, Meta's models could absorb competitive capabilities they didn't develop themselves. That's not just an intellectual property headache. It's a potential breach of service agreements with both Anthropic and OpenAI, which typically prohibit using their outputs to train competing models. Teams within Meta's Applied AI division have been told to halt tasks involving these external tools, enforce human oversight on any remaining workflows, and scrub any prior outputs from data generation processes and benchmarks. The scale of the problem was enormous Before Meta pulled the plug, usage had reached staggering levels. An internal tracking system, which Meta employees reportedly nicknamed the "Claudeonomics" dashboard, recorded 60 trillion tokens consumed within a single 30-day period. Microsoft reportedly canceled a majority of its Claude Code licenses by June 30, 2026, citing excessive token consumption as a primary driver.
Anthropic pledges $200 billion to Google Cloud, reshaping the AI infrastructure race
The five-year commitment accounts for over 40% of Alphabet's Google Cloud revenue backlog and signals a new era of hyperscale AI spending Anthropic just wrote what might be the largest check in enterprise cloud history. The Claude AI maker has committed roughly $200 billion to Google Cloud services over the next five years, a deal so large it represents more than 40% of Alphabet's entire disclosed revenue backlog for its cloud division. What the deal actually includes The commitment isn't just about renting server space. Back in October 2025, Anthropic expanded an existing agreement to acquire up to one million of Google's custom AI chips, known as Tensor Processing Units. That deal alone was valued in the tens of billions of dollars. Then in April 2026, Anthropic struck a separate alliance with Google and Broadcom targeting multiple gigawatts of next-generation TPU capacity starting in 2027. The practical result is that Anthropic's Claude models are deeply embedded in Google Cloud's Vertex AI platform, running alongside Google's own Gemini models. The two companies have been presenting jointly at Google Cloud Next conferences throughout 2025 and 2026. Anthropic isn't exclusive to Google. Claude models are also hosted on AWS and Microsoft Azure. Why this matters for the broader tech ecosystem For Alphabet investors, the math is straightforward. A $200 billion commitment over five years translates to roughly $40 billion in annual revenue from a single customer. The crypto infrastructure overlap AI training and crypto mining share a critical resource: data center capacity. Both require massive amounts of power, cooling, and physical space. As AI companies like Anthropic consume multiple gigawatts of capacity, they're competing directly with Bitcoin miners and blockchain validators for the same real estate and energy resources. Several publicly traded Bitcoin miners have pivoted toward offering AI compute services because the margins are higher and the demand is more predictable than mining rewards that fluctuate with Bitcoin's price.
Anthropic faces US regulatory scrutiny over new AI models, sparking rally in decentralized AI tokens
The Commerce Department forced Anthropic to pull its latest models offline over a jailbreak vulnerability, and crypto markets noticed immediately The US Commerce Department issued a directive on June 12 forcing Anthropic to suspend global access to its two most advanced AI models, Fable 5 and Mythos 5, after a jailbreak vulnerability raised national security red flags. The mandate required Anthropic to cut off all foreign nationals from the models, which effectively meant taking them offline entirely. The models stayed dark for roughly three weeks. Access was restored around June 30 to July 1, but only after Anthropic agreed to implement enhanced safeguards and submit to greater government oversight. Even then, the comeback was uneven: Fable 5 returned to full global access, while Mythos 5 was initially restricted to approved US organizations only. A company worth nearly $1 trillion, grounded by regulators This is a company valued at close to $1 trillion, one of the most prominent AI labs on the planet, built on a brand identity centered around safety. And it still got its models yanked offline by regulators. Just two days before the Commerce Department dropped its directive, Anthropic CEO Dario Amodei published an essay on June 10 arguing for stricter federal regulation of frontier AI. He called for rigorous testing and auditing frameworks to prevent unsafe deployments. The jailbreak vulnerability at the center of this situation involved users finding ways to bypass the safety constraints built into Fable 5 and Mythos 5. The backstory: Anthropic and the Pentagon were already at odds This wasn't Anthropic's first brush with government tension in 2026. Earlier in the year, disputes emerged regarding the military use of Anthropic's technologies, particularly involving the Department of Defense. Those disagreements raised pointed questions about whether AI systems designed with safety-first principles should be deployed in high-stakes military environments. The export control directive in June escalated that dynamic significantly. Export controls are one of the sharpest tools in the US government's regulatory toolkit, typically associated with things like advanced semiconductors and weapons systems. What this means for crypto and decentralized AI The crypto market's response was swift and predictable. Tokens associated with decentralized AI projects surged as investors drew the obvious conclusion: if a nearly $1 trillion company can have its products disabled by a single government directive, maybe there's value in systems that can't be turned off from Washington. Projects like Venice and Morpheus saw notable gains as traders rotated into assets perceived as censorship-resistant.
Anthropic details Claude Fable 5's cybersecurity safeguards and AI jailbreak framework
After a US government suspension and a narrow jailbreak scare, Anthropic rolls out a new safety classifier and proposes an industry-wide system for grading AI vulnerabilities Anthropic's newest AI model got grounded by the US government, and the company is now showing its homework on how it plans to keep things locked down going forward. Following a security incident in June 2026 that led the US Commerce Department to temporarily suspend global access to both Claude Fable 5 and its more powerful sibling Mythos 5, Anthropic has detailed a suite of cybersecurity safeguards and proposed an early framework for assessing how bad AI jailbreaks actually are. What happened, and what Anthropic built in response Fable 5 and Mythos 5 launched in early June 2026. Amazon researchers subsequently identified a narrow jailbreak vulnerability in Fable 5, one that allowed limited exploitation of known vulnerabilities. The jailbreak was characterized as non-universal, meaning it wasn't some skeleton key that unlocked everything. The US Commerce Department enforced export controls and temporarily yanked global access to both models. Access was restored on July 1, 2026, after Anthropic deployed a new safety classifier that blocks over 99% of attempts replicating the reported jailbreak. For users whose requests get flagged by this classifier, the system reroutes them to the less capable Claude Opus 4.8 model. Mythos 5 operates under stricter guardrails by default. The model is restricted to approved partners and offers greater cyber capabilities than Fable 5. Project Glasswing and the jailbreak severity framework Anthropic has been working with Amazon, Microsoft, and Google under something called Project Glasswing to propose an industry-wide system for evaluating jailbreak severity. The framework prioritizes mitigations based on two dimensions: how discoverable a jailbreak is, and how much damage it can actually cause. Anthropic has also been running its HackerOne bug bounty program with a specific focus on cyber jailbreaks targeting Fable 5. The pricing and access equation Fable 5 sits at $10 per million input tokens and $50 per million output tokens. What this means for investors and the broader market The involvement of Amazon, Microsoft, and Google in Project Glasswing is worth watching closely. When the three largest cloud platforms collaborate on safety standards, those standards tend to become the de facto industry baseline. The jailbreak severity framework, if adopted broadly, could reshape how AI companies allocate their security resources. Instead of treating every vulnerability report as a five-alarm fire, teams would triage based on structured criteria.
Anthropic's lawsuit against the Pentagon reveals deep tensions over AI safety and federal power
The AI company refused to let its models be used for autonomous weapons or mass surveillance, and the government responded by labeling it a supply chain risk. Anthropic, the AI company behind Claude, is locked in a legal battle with the US Department of Defense. The core dispute: Anthropic won't budge on two safety policies, and the Pentagon is treating the company like it's a foreign adversary because of it. The two policies in question are straightforward. Anthropic refuses to allow Claude to be deployed for lethal autonomous weapons without human oversight, and it prohibits the use of its models for mass surveillance of American citizens. But the Pentagon apparently found them unacceptable. How a $200 million deal fell apart The conflict traces back to negotiations over a potential $200 million contract between Anthropic and the DOD. Those talks collapsed when Anthropic held firm on its safety restrictions. What happened next was unprecedented. The Pentagon classified Anthropic as a "supply chain risk." This designation had historically been reserved for companies like Huawei, entities tied to foreign adversaries. Anthropic became the first US company to receive it. The Trump administration then escalated further, ordering federal agencies to stop using Claude AI entirely. On March 9, 2026, Anthropic filed lawsuits in both the Northern District of California and the D.C. Circuit, challenging the constitutionality of the government's actions. The company argued that being labeled a supply chain risk for maintaining ethical AI guidelines violated its rights and represented an abuse of federal power. Judge Rita Lin of the California court sided with Anthropic on March 26, issuing a preliminary injunction to block enforcement of certain government directives. Her language was notably pointed, suggesting the government's actions were designed to "cripple" the company and had a chilling effect on public discourse around AI safety. On April 8, the D.C. Circuit denied Anthropic's request for a stay related to the California injunction. An appeals court hearing set for May 19-20 revealed a divided panel of judges, signaling that this case is far from settled.
SpaceX shares drop $20, Elon Musk loses $110B in a day
A $20 billion bond announcement triggered a massive sell-off in SpaceX stock, vaporizing hundreds of billions in market value weeks after its blockbuster IPO SpaceX, the company that went public less than two weeks ago at $135 per share, just gave investors whiplash. A single-day plunge of more than 16% erased hundreds of billions in market capitalization and knocked over $100 billion off Elon Musk's personal fortune in one trading session. The catalyst: SpaceX announced a $20 billion bond issuance to fund AI projects and expansion plans. From IPO darling to cautionary tale in 11 days SpaceX completed its highly anticipated IPO on June 12, 2026, pricing shares at $135 and raising $75 billion. That gave the company a valuation of approximately $1.75 trillion, making it one of the largest public offerings in history. Shares surged to an intraday high of $225.64 on June 16, nearly doubling from the IPO price. Musk's net worth climbed above $1.3 trillion. On June 23, SpaceX shares cratered more than 16% in a single session. The multi-day sell-off that followed wiped out between $400 billion and $600 billion in total market value. Musk, who holds approximately 4.8 billion SpaceX shares, bore the brunt of the damage. His paper wealth dropped by $152 billion on June 23 alone. The Arkham Intelligence tracker pinned the loss at over $109 billion, though estimates vary depending on the exact timing of the snapshot. Why a bond sale spooked investors SpaceX had just completed a $75 billion equity raise. Turning around and issuing $20 billion in bonds within days signals either that the IPO proceeds weren't sufficient for the company's ambitions, or that management sees an opportunity to lock in cheap debt while investor sentiment is still warm. The bond sale was earmarked for AI initiatives and broader expansion plans. Investors, already jittery about tech valuations trading at sales multiples exceeding 100x, interpreted the move as an aggressive leverage play at exactly the wrong moment. The trillionaire threshold and what comes next Musk briefly held the distinction of being the world's first (or at least most prominent) trillionaire, with his net worth peaking above $1.3 trillion on the back of SpaceX's post-IPO rally. By early July 2026, as shares continued their descent, he had fallen below the trillionaire mark entirely. The risk for retail investors is particularly acute. Many likely bought in during the euphoric run-up to $225, attracted by the brand name and the momentum. Those buyers are now sitting on losses of 30% or more in under two weeks.
Palantir CEO criticizes OpenAI and Anthropic for overselling AI to businesses
Alex Karp called token-based pricing models 'completely, irresponsibly oversold' during a CNBC appearance, while touting Palantir's open-weight alternative Palantir Technologies CEO Alex Karp went on CNBC's Squawk Box on July 1 and did something unusual for a tech executive: he publicly torched the business models of the two most prominent AI companies on the planet. His targets were OpenAI and Anthropic, and his complaint was blunt. Their token-based pricing models, Karp argued, have been "completely, irresponsibly oversold" to enterprises that are now stuck paying escalating costs with little to show for it. Palantir's stock climbed 7.77% following the remarks. The case against token-based AI pricing Karp described a growing frustration among businesses that feel they're paying for tokens that create no value while simultaneously risking the loss of their intellectual property. Companies feed proprietary data into these models, pay handsomely for the privilege, and then worry about who actually controls what the model learned from their information. He even coined a term for the phenomenon he's criticizing: "tokenmaxxing." The word borrows the suffix from crypto and fitness culture, where it implies obsessive, often irrational maximization of a single metric. In this case, selling as many tokens as possible regardless of whether the customer is actually getting results. Palantir's alternative pitch Palantir has entered a partnership with Nvidia focused on promoting open-weight models, which are AI systems where the model weights are publicly available and enterprises can run them on their own infrastructure. The key difference from closed models like those offered by OpenAI and Anthropic: the business retains full control over its data and the model itself. Karp was notably diplomatic about at least one of his rivals. He said he appreciates private discussions with Anthropic CEO Dario Amodei, stopping short of making the critique personal. The criticism was aimed squarely at the business model, not the individual. A pattern of pointed commentary This isn't Karp's first time making waves with public statements about AI responsibility. In June 2026, he warned against executives who boast about AI-driven layoffs, arguing that such comments are both callous and strategically foolish. What this means for investors The 7.77% pop in Palantir's stock tells part of the story. Karp's remarks did not trigger any notable movements in crypto or token markets, suggesting the critique was received as a tech industry story rather than a digital asset one. The Nvidia angle is worth watching closely. The chipmaker has a financial interest in seeing open-weight models proliferate, because those models need to run on somebody's hardware. The Palantir-Nvidia partnership aligns both companies' incentives in a way that could shape how enterprise AI evolves.
SpaceX's $25B bond sale faces market skepticism despite investment-grade ratings
The largest IPO in history was followed by a massive debt offering that has already racked up $305 million in paper losses SpaceX raised roughly $75 billion in the largest IPO ever recorded on June 12. Ten days later, it came back asking for $25 billion more, this time in bonds. The market said yes, then immediately started having second thoughts. The senior unsecured bond offering, announced on June 22, was technically oversubscribed. Orders piled up to approximately $89 billion, more than three times the deal size, which had already been upsized from an original $20 billion target. But oversubscription and enthusiasm are not the same thing, and what happened next made that distinction painfully clear. The numbers tell a complicated story SpaceX's bonds received investment-grade ratings from all three major agencies: Moody's assigned Baa1 with a stable outlook, S&P gave it a BBB, and Fitch followed suit. The bonds priced approximately 0.4 points wide of BBB averages. The five tranches mature between 2031 and 2056, with yields ranging from 5.35% to 6.65%. Shortly after pricing, the bonds recorded around $305 million in paper losses as secondary trading revealed sluggish demand. The proceeds are earmarked for repaying a bridge loan and funding general corporate purposes, including artificial intelligence initiatives through SpaceX's xAI unit. From rocket fuel to market turbulence SpaceX's IPO priced shares at $135, giving the company an initial market capitalization exceeding $1.7 trillion. It began trading on Nasdaq under the ticker SPCX. SpaceX's stock dropped between 16% and 23% in the days following the listing, erasing over $600 billion in market value. Rating agencies pointed to Starlink's recurring revenue base and SpaceX's reusable rocket technology as core strengths supporting their investment-grade assessments. The company also disclosed holding 18,712 BTC on its corporate treasury as of March 31, 2026, valued between $1.2 billion and $1.3 billion at the time. What this means for investors For investors holding both SpaceX equity and debt, the concentration risk is worth noting carefully. A stock that can shed $600 billion in days is volatile enough on its own. Pairing that with bonds already trading below their issue price creates a correlated downside scenario. The $305 million in paper losses may seem modest relative to a $25 billion offering, roughly 1.2%. But in the investment-grade bond market, where stability is the entire selling proposition, early losses send a signal that lingers.
SpaceX joins the Nasdaq-100 on July 7, boosting index inflows by billions
Passive fund managers must buy SPCX shares after Nasdaq's rule change fast-tracks the aerospace giant into the index just weeks after its IPO. SpaceX went public on June 12, 2026, under the ticker SPCX on Nasdaq. Less than a month later, it is joining one of the most tracked indexes on Earth. The company is set to enter the Nasdaq-100 on July 7, 2026, a timeline that would have been impossible under the old rules. Nasdaq updated its index inclusion criteria in May 2026, creating a fast-entry provision for large new listings that skips the traditional minimum float waiting period. SpaceX is among the first major beneficiaries of that change. The mechanics of forced buying When SpaceX joins the Nasdaq-100, every ETF and passive mutual fund benchmarked to that index has to buy shares to stay in balance. J.P. Morgan estimates passive fund inflows tied to this inclusion could reach approximately $4.3 billion. Other projections put the figure higher, somewhere between $7 billion and $10 billion, depending on how SpaceX's float and weighting shake out within the index. SpaceX's initial public float at listing was roughly 4%, meaning the vast majority of shares remain locked up, primarily with Elon Musk and other major insiders. As that float expands over time, the index weighting and the corresponding passive demand could shift materially upward. What history says about index additions SpaceX enters the Nasdaq-100 with a post-IPO market capitalization in the trillions, putting it among the largest publicly traded companies in the world almost immediately after listing. The company had already entered the Russell 1000 Index under fast-entry provisions, so this is not its first experience with the index inclusion playbook. The Nasdaq-100 addition, though, carries considerably more passive AUM behind it. SpaceX's financials reflect the profile of a high-growth company still investing aggressively. Revenue has grown substantially, but the company carries GAAP losses typical of firms prioritizing long-term infrastructure over near-term profitability. What this means for investors The float constraint is the wild card. With roughly 4% of shares publicly available at IPO, liquidity is tight relative to the size of the passive demand estimates. As lock-up periods expire and more shares enter circulation, the index weighting recalibrates and the demand picture evolves. Nasdaq's decision to update its inclusion rules in May 2026 also sets a broader precedent. If large, newly public companies can now enter the Nasdaq-100 within weeks of their IPO, future mega-listings will carry a built-in catalyst in the form of near-immediate index inclusion demand.
Kalshi and Polymarket see 75% growth in June trading volume to $45B
Prediction markets are scaling at a pace that makes most DeFi protocols look like they're standing still, with sports betting driving the bulk of the action. Prediction markets just had their biggest month ever. Kalshi and Polymarket processed nearly $45 billion in combined trading volume during June, a 75% jump from the prior month that cements the sector as one of the fastest-growing corners of the financial landscape. To put that number in perspective, combined monthly volumes for the two platforms sat below $5 billion as recently as September 2025. Nine months later, they've grown roughly ninefold. Kalshi is running away with market share Of the two platforms, Kalshi is doing most of the heavy lifting. The US-regulated exchange processed approximately $22.6 billion in notional volume in June, significantly outpacing Polymarket and commanding about 74% of weekly market share in recent tallies. The secret ingredient is not exactly a secret: sports betting. An estimated 80% of Kalshi's trading volume comes from sports markets, turning what was originally pitched as a platform for hedging real-world events into something that looks a lot more like a sportsbook with extra steps. Polymarket takes a different approach. Sports betting accounts for a smaller share of its volume, somewhere in the range of 39% to 45%, with the platform maintaining broader interest in political outcomes and crypto-adjacent markets. Weekly trading volumes across both platforms recently hit $13.9 billion. The trajectory from under $5 billion monthly in September to roughly $24 billion in April to $45 billion in June suggests the growth is accelerating, not plateauing. Big valuations, bigger expectations The venture capital world has noticed. Kalshi closed a funding round in May 2026 that valued the company at $22 billion, with some projections suggesting a potential path to $40 billion. Polymarket's valuation sits around $15 billion, built on blockchain infrastructure. Neither platform has prominently tied any crypto tokens to their core operations, despite Polymarket being built on blockchain rails. What this means for investors and traders The dominance of sports betting also creates concentration risk. If regulators decide to draw sharper lines between prediction markets and traditional sports gambling, Kalshi's volume engine could face headwinds. Regulatory scrutiny around prediction markets has been increasing, particularly regarding potential insider trading on geopolitical and political events. For the competitive landscape, the divergence between Kalshi and Polymarket is worth watching. Kalshi's bet on sports volume is paying off spectacularly right now, but Polymarket's more diversified approach could prove more resilient if regulatory winds shift. A platform that derives 80% of its volume from one category is inherently more vulnerable to category-specific disruption than one with a broader base.
SpaceX showcases prototype device to transform human-AI interaction
A slim AI-powered handset was reportedly shown to SpaceX investors, though Elon Musk has already denied its existence SpaceX apparently wants to put AI in your pocket, and not through an app store. The company reportedly showed investors a prototype device, described as a slim handset thinner than an iPhone, designed to fundamentally change how people interact with artificial intelligence. The catch: Elon Musk says it doesn't exist. The Wall Street Journal reported on July 1, 2026, that SpaceX presented the prototype to investors, integrating AI technology derived from xAI, the company SpaceX acquired back in February 2026. The device reportedly runs on Qualcomm chipsets with a proprietary operating system, and leverages Grok models to deliver what's being described as a next-generation AI interaction experience. Musk responded on X by calling the report "utterly false." What we know about the device The device is distinct from Neuralink, Musk's brain-computer interface venture. Where Neuralink aims to merge human cognition with computing at the neural level, this prototype, if real, would function as a consumer hardware product. The xAI connection and SpaceX's expanding ambitions SpaceX's acquisition of xAI in February 2026 raised eyebrows at the time. This prototype, if it exists, would retroactively make that acquisition look strategic rather than impulsive. xAI's Grok models were already competitive with offerings from OpenAI and Anthropic. There's also the Starlink angle. SpaceX already operates the world's largest satellite internet constellation, and has been forging partnerships to bring satellite connectivity to consumer mobile devices. A SpaceX-built handset could theoretically ship with native Starlink connectivity, giving it coverage in areas where traditional cellular networks don't reach. What this means for investors Musk's denial creates an unusual situation. The report is specific enough to move markets, but the company's CEO has explicitly rejected it. Investors are left choosing between a major financial publication's sourcing and the word of a CEO who has, historically, had a complicated relationship with public statements about his companies' plans. What's clear is that the mere possibility of SpaceX entering the consumer device market has implications for several sectors. Qualcomm stands to benefit if the partnership is real. Traditional smartphone manufacturers would face a new competitor with deep pockets and a fanatical customer base. Telecom companies that have built their business models around terrestrial infrastructure might need to reconsider their competitive positioning against a company that can deliver connectivity from orbit.
