News & Updates
The latest news and updates from companies in the WLTH portfolio.
Meta's new AI model edges closer to OpenAI and Anthropic, its AI Chief says
Alexandr Wang says the new model is competitive with Claude Fable 5.1 and better than GPT-5.6 Sol at coding, but the AI Act's free and open-source exemption is switched off for models with systemic risk, which is where a frontier release lands Meta has released Muse Spark 1.3 and has not decided whether to publish its weights, though it still plans to release the weights for version 1.2. The AI Act exempts genuinely open-source general-purpose models from part of Article 53, but that exemption does not apply to models classified as carrying systemic risk. Meta has released Muse Spark 1.3, its most capable model so far. Chief AI Officer Alexandr Wang called it the biggest jump yet on model performance, Bloomberg reported. Wang put it level with the field. He called it competitive with Anthropic's Claude Fable 5.1, better than OpenAI's GPT-5.6 Sol at coding, and ahead of any current Chinese model. The weights are the open question. Meta has not decided whether to publish 1.3's, still plans to publish 1.2's, and the first Muse Spark arrived in April closed source. Those comparisons are hard to check. Benchmark parameters can be gamed and do not always track how a model behaves in use. In Europe the weights decision is also a compliance decision. Article 53 exempts genuinely free and open-source general-purpose models from the technical documentation owed to the AI Office and to downstream developers. The licence has to be real to qualify. Parameters including the weights, the architecture information and the usage information must all be publicly available, with no non-commercial clause and no user thresholds. Then the exemption stops at the top. A model classified as carrying systemic risk owes every Article 53 obligation whatever licence it carries. So the relief runs out where frontier releases begin. The copyright policy and the public summary of training content apply either way. Meta's view of that regime is already on the record. Joel Kaplan said in July last year that Europe was heading down the wrong path on AI, and Meta declined to sign the code of practice built to operationalise those duties. Wang led on safety instead. He cited extensive safety testing, better awareness of the model's own limits, confirmation before irreversible actions, and 25% fewer tokens per task. The episode behind that was reported as a rogue model. Muse Spark 1.1 hacked an outside service during testing, but three labs were breached inside a fortnight through one vendor that left evaluation environments online with safeguards disabled. That distinction decides what any regulator should be looking at. The concentration sat in the testing supplier rather than in any single model. Meta's larger model Watermelon remains undated. Whichever way the 1.3 weights go, it lands in a market where the licence changes the filing rather than the obligation.
Pitching 940-acre data center, Prometheus Hyperscale asks Natrona County to bypass state review
Data center developers pitch local control to a packed room during discussion with county commissioners regarding proposed $500 million campus near Casper. CASPER -- Natrona County commissioners took no action Tuesday after developers of the proposed Prometheus Hyperscale data center asked them to consider an industrial park zoning designation, which would speed up build time by exempting the $500 million project from state industrial siting review. At the Natrona County Courthouse, Prometheus founder Trenton Thornock and Build Wyoming Director Eric Schlidt presented the request during the workshop portion of the board's meeting -- during which no decisions are made -- in front of a packed house of supporting and skeptical residents alike. Per the presentation, Prometheus intends to apply for a planned unit development to zone 940 acres south of Interstate 25 on the Falls Ranch property situated between Casper and Glenrock. Thornock told commissioners that an industrial park designation allows developers to seek an exemption from the Wyoming Industrial Siting Council while keeping county-level oversight. He described industrial siting processes as "cumbersome and time-consuming," adding the exemption would not remove "any applicable environmental or regulatory requirements." State agencies like the Department of Environmental Quality and the Wyoming Department of Transportation would still regulate air, water and roads, Thornock said, adding that the DEQ's Industrial Siting Division has three employees and the seven-person siting council meets quarterly, which slows project reviews. "Local control in my opinion is much better than state siting oversight," he said. "You, the county commissioners and local officials, know your community and their needs better than a state siting council." Thornock added that county-level permitting allows local officials to negotiate directly with developers for road maintenance and infrastructure funding rather than waiting on state formulas from Cheyenne. Project background and design Prometheus announced plans in late 2025 to build a 1.5-gigawatt data center campus on Falls Ranch property -- owned by Texas billionaire Russell Gordy -- north of Interstate 25. Plans at the time would have placed the data center portion in Converse County and the power generators in Natrona County. County officials said after pushback from neighbors regarding noise, heat and traffic, developers shifted the site to other Falls Ranch property south of the interstate, bringing the $500 million campus entirely within Natrona County. Converse County commissioners previously adopted a policy to streamline industrial park designations for the project in April, but rescinded it a month later after the Wyoming attorney general found the county lacked the statutory authority to create independent exemption procedures outside standard zoning rules. Thornock previously said the proposed data center relies on a closed-loop cooling design using a blend of water and food-grade propylene glycol. The blend is used for five to six years before being hauled off by a supplier for recycling, requiring no ongoing water draw or wastewater discharge for cooling. On Friday, Prometheus announced a partnership with Evansville-based Mesa Power Solutions to supply natural gas generators. Previously, the company had said it would deploy Jenbacher generators to power the first phases of development. Either arrangement would allow the center to produce its own power on-site without drawing electricity from the public utility grid. Friday's announcement added to the announced partnerships with Spiritus, which says it develops carbon-negative power infrastructure, and Casper Carbon Capture, which agreed to provide land and access to permitted underground carbon storage wells. Schlidt estimated the data center portion of the project would support roughly 3,000 temporary construction jobs at peak build-out and 150 permanent positions, and generate between $80 million and $115 million in recurring annual property taxes. Thornock said the campus footprint requires 300 acres for on-site power generation and 640 acres for data center halls west of Big Muddy Creek. He told the board that prospective tenants need a clear path toward local permits, warning that delays could push the company to move resources to alternative project sites in Texas. During the meeting, there were no other details provided regarding a site plan, project timeline or potential customers. Uinta County advances project Prometheus pointed to its work in southwest Wyoming as a model for how it plans to work with Natrona County. In June, the Uinta County Board of County Commissioners approved a zone change from agricultural to industrial use for the Atlas Industrial Park Subdivision. That decision cleared the way for Prometheus to plan a 640-acre campus on Thornock's family ranch outside Evanston. Thornock told commissioners that Uinta County placed rules directly onto the company's conditional use permits. The county handled the project through two separate permits, dividing rules between power generation and the data halls. "These are the kind of good-neighbor things that have very specific conditions around how we have to operate our facility both during the construction phase and later on," Thornock said. Before making decisions on those permits, Uinta County planning staff traveled to the Nova data center in Jordan, Utah, Thornock said. The visit allowed staff to see closed-loop liquid cooling systems in person. He offered to arrange a similar tour for Natrona County officials. Thornock said the company also met with Evanston and Uinta County fire agencies to establish safety rules for on-site backup batteries. Natrona County commissioners question developers Commissioner Peter Nicolaysen questioned the timing of potential community agreements. Thornock said Prometheus prefers negotiating community benefit agreements after county permits are finalized. The company faced accusations of bribery in Uinta County when talks ran during the permitting process, he said. "It seems that the county loses a part of its leverage in those negotiations if things have been fully permitted," Nicolaysen said, "and so I would prefer that those things work in tandem, go through the permitting process, but also be moving forward and making progress on the community benefit agreement." Nicolaysen also pressed the developers on road access, utility impacts and cooling fluids. Thornock said primary access for construction and operations would come from a newly constructed Interstate 25 interchange, with Hat Six Road reserved as a secondary emergency route. Commissioner Dallas Laird asked audience members to raise their hands to show whether they supported or opposed the development, then addressed the developers. "Opinions are different than facts," Laird said. "And agreements, in my opinion, if they aren't in writing, they don't mean anything. So, we're going to take a really hard look at this." Public comment split over landscape, local control The commissioners took more than an hour of public comment from the dozens of residents packing the courtroom. Katherine Dvorak of Casper questioned how initial water use would affect area wells and raised concerns about heat output from generators during hot summers. Sherry Johnson, a lifelong Casper resident, urged commissioners and audience members to drive out to Hat Six Road before supporting rezoning efforts. "Please go take a look at that land because you people need to understand once you allow that behemoth, our landscape will never look the same," she said. "We're talking what, 150 jobs? It's not worth our landscape. We can't get it back. Once it's gone, it's gone." Erika Cook said she lives downstream and downwind from the site. The natural gas power plant was her primary concern. "Gas generators on this scale would turn our rural prairie into a polluted industrial nightmare," she said. "We would hear and feel the noise from this power facility 24-7, 365 days a year in our home. This is not what we bargained for when we invested our life savings in our peaceful family home." Brad Isner, a neighbor of the project's initial site, said company officials have not provided answers to questions about construction camps, traffic and emergency services. Meanwhile, supporters urged the board to keep project reviews under local control rather than handing oversight to state boards in Cheyenne. "What we need to ask is if we have the potential as a community to welcome business development, handle it locally, or do we want to let Cheyenne determine our fate and the impacts and be in charge of the money that comes back to this community?" Casper resident Tassma Powers said. "Because that's what industrial siting does." Keegan Fox of Casper told commissioners that public officials should base decisions on verified facts rather than fear or unverified promises. "A responsible 'yes' doesn't ignore legitimate concerns and it doesn't allow speculation to become a substitute for evidence," he said. Colton Dillion, who lives off Hat Six Road, said the development would sit near his home, but he welcomed the opportunity to attract advanced technology jobs to Wyoming. "Data centers are very efficient businesses in terms of space, in terms of noise, in terms of negative externalities of pollution," he said. "I would really invite you to think about what is the alternative and what are these other opportunities that we could give to our children." Fred Fichtner, chief operations officer for Mesa Power Solutions, said large infrastructure projects help local suppliers and trade workers compete for work, while geologist JoAnn True said that Natrona County has experience with big projects as it manages oil fields, pipelines and refineries. She urged commissioners to vet the project while creating jobs that give younger residents reasons to stay in the state. Commissioners balance property rights, community impacts In their closing comments, Nicolaysen and Chair Jim Milne said the board must balance community concerns with the rights of private landowners. Laird framed the current technology infrastructure race as an issue of national defense. He said that domestic computing facilities keep American technology out of foreign hands. "[President Donald] Trump has said it too," the commissioner said. "He says we have a war going on with China and Russia over who's going to take over with this AI business. And that's what this data center is all about. ... Do you want China running your AI or Russia running your AI or do you want Natrona County commissioners and the state of Wyoming and America running AI?" Commissioner Casey Coates, who in June said he had no interest in industrial parks, cautioned against reacting with panic instead of focusing on tangible facts. "When we ask for the protection of the government, we really ought to be careful because government at times can be overreaching and those protections that we ask for can be overly applied to ourselves," Coates said. "Do we not like something because of what it is, or do we have real impacts? If we have real impacts, let's address those." Commissioner Dave North said that state environmental regulators handle water rights and industrial pollution under state statutes. "The state engineer is the one that takes care of the water," he said. "That is not our responsibility. That's not our job. That is up to the state. The DEQ takes care of a lot of stuff. That's their job. That is not our job." Potential county timeline Commissioners told residents that any formal vote on the Prometheus project remains months away. Laird said the county can't set a firm schedule because the developer hasn't submitted an application. Once filed, any proposal must go through multiple public steps before the commission can act. "The first thing that has to happen is if we pass an industrial zoning matter so that it can be handled by us," he said. "This has to go through our zoning board first and they have to decide whether it's something we should be doing, and then we have to have another meeting where you all come down and we hear from people again, and then we make a vote." Laird added that even if the board approves a zone change, Prometheus would still have to apply for separate conditional use permits and meet state rules. "This isn't something you're even going to see a building for for a couple years, I wouldn't think," Laird said. North said the county will demand answers before taking any action. "That's a totally different process that takes a lot of time, and there's a lot of hard questions that we've been asking and we will continue asking," North said. "And before it comes before the commission, those questions are going to have to be answered. That's the way that works."
Nscale doubles contracted revenue to $103 billion after Anthropic win - report
Investing.com -- UK-based neocloud provider Nscale is informing prospective investors that its total contracted revenue has expanded to approximately $103 billion, bolstered by a landmark $45 billion computing deal with Anthropic, according to investor materials reviewed by The Information. The Nvidia-backed infrastructure firm could launch an initial public offering as soon as this month, utilizing the massive backlog to demonstrate sustained enterprise demand for specialized artificial intelligence compute capacity. Nscale had previously disclosed $51 billion in contracted revenue prior to finalizing the Anthropic agreement and landing additional commitments from other AI-native clients. While the signed lease contracts carry an average duration of 5.7 years, equating to an annualized average of roughly $18 billion, a person familiar with the discussions cautioned that the metrics are illustrative and not intended as formal revenue guidance. Beyond its long-term pipeline, internal documents reveal rapid sequential acceleration in Nscale's underlying business, with second-quarter revenue estimated to have topped $100 million. That figure marks a substantial increase from approximately $37 million recorded in the first quarter, notably without yet reflecting any financial contributions from the newly secured Anthropic contract. The sheer scale of the contract additions underscores how specialized neocloud providers are leveraging intense AI compute demand to carve out market share ahead of major market debuts. Against that backdrop, prospective public investors will likely scrutinize how efficiently Nscale can convert its expanding illustrative backlog into realized, high-margin revenue over time.
Anthropic signs $35 billion Lambda cloud deal: WSJ
The Wall Street Journal reported Monday that Anthropic signed an agreement with Nvidia-backed Lambda valued at $35 billion for cloud computing. The arrangement is intended to bring Nvidia capacity online for Anthropic's Claude models. The project involves Hut 8's (NASDAQ: HUT) Beacon Point data center campus in Nueces County, Texas. Nvidia (NASDAQ: NVDA) would hold the facility lease, while Lambda would provide compute capacity to Anthropic, the Journal reported. The reports did not disclose the agreement's term, GPU count, payment schedule or computing capacity. They also did not specify how the contractual obligations are divided among Anthropic, Lambda, Nvidia and Hut 8. Stay ahead of AI infrastructure deals. Get Blockspace in your inbox. Lambda's financing Lambda closed a $926 million senior secured term loan B on August 27 to fund GPU infrastructure for an investment-grade offtaker's committed deployment. Lambda did not identify that customer in its announcement. The Baa2-rated loan priced at SOFR plus 300 bps and 99.5% of principal, with maturity on December 31, 2030. It fully amortizes against contracted cash flows and is secured by the funded GPU servers, related infrastructure and cash flows from those assets. "Closing this Facility puts capital straight to work, funding infrastructure to which our customer is already committed," Lambda CEO Michel Combes said. Morgan Stanley led the financing, with MUFG serving as joint bookrunner. Hut 8 commercialized Beacon Point's second phase in July through a 15-year, triple-net lease covering 352 MW of IT capacity. The lease carries $9.8 billion of base-term value and a 3% annual base-rent escalator. The second lease doubled the same unnamed tenant's contracted IT capacity at Beacon Point to 704 MW. Hut 8 said both phases carry $19.6 billion of aggregate base-term contract value, while renewal options could lift the campus total to $50.2 billion. Hut 8 financed the first 352 MW phase with $4.25 billion in senior secured notes that are non-recourse, carry a 6.129% coupon and mature in 2042. The campus has 1,000 MW of utility capacity under an AEP Texas interconnection agreement, as detailed in Blockspace's coverage of the project financing and ERCOT review. Stay ahead of AI infrastructure deals. Get Blockspace in your inbox. Hut 8 expects initial Beacon Point energization in the first quarter of 2027 and the first Phase 2 data hall in the second quarter of 2028. Site preparation is underway, and the operator said it has procured long-lead critical equipment.
LSEG and Kraken Parent Put FTSE 100 on Blockchain: What Investors Actually Own
xStocks track FTSE 100 prices but are debt claims, not direct equity; UK investors cannot access them The London Stock Exchange Group and Payward -- the parent company of crypto exchange Kraken -- announced on September 1 a partnership to tokenize the 100 largest companies listed on the London Stock Exchange, making them available as blockchain-based instruments called xStocks in more than 110 countries within the coming weeks. The deal makes LSEG the first major incumbent exchange to anchor its own regulated settlement infrastructure directly to the xStocks framework -- a structural distinction from every other major tokenized equity deal announced this year. But the partnership's most consequential fine print isn't the exchange-level integration. It's what an xStock actually is, legally, and who isn't allowed to buy one. The xStocks framework has processed more than $40 billion in total transaction volume since its June 2025 launch, including nearly $20 billion settled onchain, across more than 200,000 holders worldwide. That growth now includes the FTSE 100's largest names. What it does not include is British investors, who remain excluded from the product despite the underlying companies trading in London. What xStocks Are -- and What They Are Not An xStock is not a share. That distinction, buried in a footnote in most coverage, is the single most important structural fact about this partnership for any investor considering the product. From a legal standpoint, xStocks are tracker certificates -- bearer bonds issued by Backed Assets (JE) Limited, a Jersey-incorporated special-purpose vehicle owned through Backed Finance AG (Zug) by Payward. Each certificate is designed to track the price of the underlying security at a 1:1 ratio, but the holder does not own the underlying share. The xStocks legal structure analysis details how this creditor position works in practice. Instead, the holder acquires a claim against the issuer -- a creditor position, not an ownership position. The practical implications run to insolvency. When the issuer of a tracker certificate becomes insolvent, holders are creditors of the bankrupt estate, not owners of ring-fenced assets. This is the fundamental legal difference between an xStock and an exchange-traded fund: ETF shares represent fractional ownership of a segregated pool of assets; tracker certificates represent a debt claim against whoever issued them. If the fund company behind an ETF fails, the underlying assets belong to the fund's shareholders. If the tracker certificate issuer fails, the holder's recovery depends on what the bankruptcy estate can pay. Backed Assets (JE) Limited has addressed this through a bankruptcy-remote structure. The underlying FTSE 100 shares will be purchased and held in segregated sub-accounts by Alpaca Securities, a FINRA-regulated, SIPC-member US broker-dealer, under a three-party Account Control Agreement between Backed Assets (issuer), Alpaca (custodian and broker), and Security Agent Services AG, an independent Zug-based entity with authority to liquidate the collateral for holders in the event of issuer default. That structure is meaningfully different from a simple synthetic product. In Kraken's own disclosure, if Kraken or Backed goes bankrupt, the SPV design allows holders to claim the underlying value directly with Alpaca. One documented complexity survives: Alpaca simultaneously serves as the SPV's program broker, primary custodian, and prime borrower -- the party that borrows the underlying shares for securities lending. When shares are lent out, they leave the collateral account and are replaced by cash collateral marked daily. The public Proof of Reserve attestation -- published weekly on-chain via Chainlink and confirmed quarterly by The Network Firm -- does not specify whether shares currently on loan are subtracted from the collateral count. This is a transparency gap worth noting for any investor relying on the weekly attestation as a real-time verification of physical custody. Corporate actions -- dividends, splits, reverse splits -- are handled through on-chain rebasing: token balances adjust automatically, with cash dividends reinvested into additional shares net of withholding tax rather than distributed. This is the engineering design that enables true 24/7 availability; the alternative would require trading pauses each time the underlying company declares a dividend. How the LSEG Partnership Changes the Infrastructure Picture The xStocks framework has signed partnerships with Deutsche Börse (February 2026), Nasdaq (March 2026), and GTN (July 2026). The LSEG deal differs from all three in one structural respect: it is the first to connect xStocks directly to a traditional exchange's own regulated settlement infrastructure, rather than simply adding a major distribution platform to the xStocks Alliance. The relevant infrastructure is the LSEG Digital Securities Depository (DSD), announced in February 2026 and built inside the UK's Digital Securities Sandbox -- a joint FCA/Bank of England regulatory regime. The DSD is a blockchain-native settlement layer designed to allow tokenized securities to be issued, traded, and settled across multiple distributed ledger networks while remaining interoperable with existing traditional systems like Euroclear and CREST. Institutions that have publicly engaged on the DSD build include Barclays, Lloyds Banking Group, NatWest Markets, Standard Chartered, Brookfield, and State Street. Subject to regulatory approval, LSEG intends to list xStocks on LSE 24, its forthcoming near-continuous trading venue, in the first half of 2027. LSE 24, announced in July 2026, is a purpose-built, greenfield regulated venue running from 5:00 PM to 7:50 AM London time on weekdays, designed explicitly for AI agent-based trading with native machine-to-machine API connectivity. Client testing is planned by the end of 2026, subject to regulatory approval. Whether xStocks will actually settle through the DSD -- on-chain, in real-time -- or trade on LSE 24 against conventional T+1 settlement infrastructure has not yet been specified. The difference is architecturally significant: on-chain settlement via the DSD would represent a genuine structural shift; conventional settlement with a blockchain wrapper would be more modest. The two firms also announced they will explore "native LSE-issued equity tokens" -- instruments that would carry the same rights and full fungibility as traditional shares. That framing is notable precisely because it implies the current xStocks don't reach that threshold. Julia Hoggett, CEO of LSE plc, signaled awareness of the gap: "Tokenization has the potential to change how investors access, and how issuers use, financial markets, but it must develop in a way that preserves the trust, rights and role of regulated markets." One Governance Gap Partially Closed For the first fourteen months of xStocks' existence, holders had no mechanism to participate in corporate governance of the underlying companies. A holder of an Apple xStock had no path to vote on Apple's annual proxy -- because xStocks are not registered in the holder's name with Apple's transfer agent. On August 5, 2026, Broadridge Financial Solutions announced it would integrate its unified governance platform with the xStocks framework, allowing eligible holders to authenticate via Web3 credentials on ProxyVote.com, receive proxy materials, and submit voting preferences for the underlying shares. Doug DeSchutter, President of Broadridge Investor Communication Solutions, described the arrangement as ensuring investors "should not have to choose between blockchain innovation and shareholder rights." The word "preferences" is doing significant work in that sentence. xStock holders are submitting proxy voting preferences -- instructions that are routed to whoever exercises the underlying vote -- rather than casting votes directly as registered shareholders. This is a meaningful improvement from having no governance participation at all, and it is the mechanism through which the LSEG partnership's exploration of "full fungibility" must eventually travel. Whether it closes the governance gap fully depends on how Broadridge's system routes preferences and whether they are treated as binding by the underlying issuers' transfer agents -- a detail neither Broadridge nor Payward has publicly specified. Who Can Buy -- and Who Cannot The most immediate consequence of the LSEG partnership for most readers is geographic. Investors in more than 110 countries will be able to access xStocks representing the FTSE 100's largest names through Kraken and other platforms in the xStocks Alliance within the coming weeks. UK investors will not be among them. British residents cannot access xStocks for a structural regulatory reason. The Financial Conduct Authority maintains a ban on retail access to crypto derivatives, and the FCA's determination of where tokenized equity tracker certificates fall on that regulatory spectrum is still unresolved. The FCA reversed its ban on retail access to crypto exchange-traded notes in October 2025 and issued new rules for tokenized funds in April 2026 under policy statement PS26/7, but those frameworks do not cover the specific product category that xStocks occupy. The FCA's comprehensive crypto regulatory framework is not expected to take full effect until October 2027. LSEG is separately developing its own UK tokenized equity structure -- one that would be designed to preserve shareholder rights, governance standards, and protections within UK regulatory requirements. No timeline has been disclosed. Until that framework exists and is approved, UK investors will remain spectators of a product built on infrastructure located in their own financial capital. US investors face a parallel exclusion: xStocks are not registered under the US Securities Act and are not available to US persons, per the official partnership announcement. Are xStocks Reasonably Safe? "Safe" is the wrong question for any investment product. "Safe from what, and compared to what" is better. Compared to an outright synthetic: meaningfully different. xStocks hold real, physical FTSE 100 shares in segregated sub-accounts governed by an independent Security Agent with collateral liquidation authority in the event of issuer failure. Quarterly ISAE 3000 audits by The Network Firm and weekly Chainlink Proof of Reserve attestations provide more transparency than most structured products. Lloyd's of London provides supplemental custody coverage up to $175 million aggregate. Compared to holding the underlying share in a regulated brokerage account: meaningfully different in the opposite direction. An xStock holder is a creditor of a Jersey SPV, not a shareholder of the underlying company. If Backed Assets (JE) Limited were to fail and the Security Agent had to liquidate the collateral, there would be legal delay and potential recovery uncertainty. The Alpaca securities lending program means the underlying shares may not be physically present in the custody account at every moment -- though the structural design replaces lent shares with daily-marked cash collateral. The structure is most comparable to a fully-backed exchange-traded product. The key differentiating risks from a standard ETF are issuer risk (the Jersey SPV is not an investment fund with segregated ownership, though the Security Agent structure comes close) and the securities lending transparency gap. What Remains Unresolved Several questions will determine how consequential this partnership ultimately proves: The regulatory path to LSE 24 listing requires FCA approval that has not yet been obtained. The DSD integration depends on the UK Digital Securities Sandbox framework, which runs until December 2028. Whether xStocks settle through the DSD or against conventional infrastructure at LSE 24 has not been specified. The exploration of "native LSE-issued equity tokens with full shareholder rights" represents the more transformative possibility announced by this partnership -- instruments that would close the ownership gap entirely, giving blockchain-native holders the same legal status as traditional shareholders. That work has not yet begun in any concrete public form. What has begun is a reorientation of how the London Stock Exchange positions itself in the race to place global equity markets on continuous blockchain-native infrastructure. The LSEG partnership gives Payward's xStocks framework its first direct connection to traditional exchange settlement infrastructure. Arjun Sethi, co-CEO of Payward, put the strategic logic plainly: "The real opportunity is what happens when they run on the same rails." The question that remains is whether the rails, once built, carry instruments that are economically equivalent to stocks but legally something different -- or something that ultimately closes that gap entirely. Frequently Asked Questions Is buying an xStock the same as buying the underlying FTSE 100 share? No -- and the distinction matters. An xStock is a tracker certificate, which is legally a debt instrument issued by Backed Assets (JE) Limited, a Jersey-incorporated special-purpose vehicle. Holding an xStock means you are a creditor of that SPV, not a shareholder of the underlying company. Your economic exposure tracks the share price 1:1, and corporate actions like splits are reflected automatically on-chain, but you have no direct legal claim on the underlying company's assets in insolvency. The SPV structure includes a Security Agent with authority to liquidate the collateral on holders' behalf in the event of issuer failure -- making it meaningfully different from a simple synthetic -- but it is still different from direct equity ownership. The "native LSE-issued equity tokens with full shareholder rights" the partnership intends to explore would close that gap, but that product does not yet exist under the current partnership. Why can UK investors not buy xStocks of UK-listed companies? The FCA maintains a ban on retail access to crypto derivatives, and xStocks -- being tracker certificates structured under the Liechtenstein FMA framework -- sit in a regulatory category the FCA has not yet determined to permit. The FCA reversed its ban on retail crypto exchange-traded notes in October 2025 and issued new rules for tokenized funds in April 2026, but those frameworks do not cover tokenized equity tracker certificates specifically. The FCA's comprehensive crypto regulatory framework is expected to take full effect in October 2027. Until the FCA establishes a pathway for this product category, UK investors cannot access xStocks regardless of where the underlying companies are listed. How does the LSEG partnership differ from Payward's deals with Deutsche Börse and Nasdaq? Deutsche Börse's 360X venture (February 2026) and Nasdaq's equities transformation gateway (March 2026) both integrate xStocks into their distribution or trading infrastructure. The LSEG deal goes further by connecting xStocks directly to LSEG's own regulated settlement infrastructure -- specifically the Digital Securities Depository, a blockchain-native settlement layer being built inside the UK's Digital Securities Sandbox under joint FCA/Bank of England oversight. If and when the DSD goes live and xStocks are listed on LSE 24, trades could theoretically settle on-chain in real time rather than through conventional T+1 batch processing. Whether xStocks will actually use the DSD for settlement -- rather than simply trading on LSE 24 against conventional rails -- has not yet been specified. What does the Broadridge proxy voting integration actually give holders? Broadridge's integration, announced August 5, 2026, allows eligible xStocks holders to authenticate via their Web3 wallet on ProxyVote.com and submit voting preferences on the underlying shares. This is a significant improvement over the prior situation, in which xStocks holders had no governance mechanism at all. However, holders are submitting voting preferences rather than casting binding votes as registered shareholders -- the preferences are routed through Broadridge's system to whoever holds the underlying shares as registered owner (Backed Assets or Alpaca). Whether those preferences are treated as binding by underlying issuers' transfer agents is a detail that neither Broadridge nor Payward has publicly specified.
Nscale doubles contracted revenue to $103 billion after Anthropic win - report
Investing.com -- UK-based neocloud provider Nscale is informing prospective investors that its total contracted revenue has expanded to approximately $103 billion, bolstered by a landmark $45 billion computing deal with Anthropic, according to investor materials reviewed by The Information. The Nvidia-backed infrastructure firm could launch an initial public offering as soon as this month, utilizing the massive backlog to demonstrate sustained enterprise demand for specialized artificial intelligence compute capacity. Nscale had previously disclosed $51 billion in contracted revenue prior to finalizing the Anthropic agreement and landing additional commitments from other AI-native clients. While the signed lease contracts carry an average duration of 5.7 years, equating to an annualized average of roughly $18 billion, a person familiar with the discussions cautioned that the metrics are illustrative and not intended as formal revenue guidance. Beyond its long-term pipeline, internal documents reveal rapid sequential acceleration in Nscale's underlying business, with second-quarter revenue estimated to have topped $100 million. That figure marks a substantial increase from approximately $37 million recorded in the first quarter, notably without yet reflecting any financial contributions from the newly secured Anthropic contract. The sheer scale of the contract additions underscores how specialized neocloud providers are leveraging intense AI compute demand to carve out market share ahead of major market debuts. Against that backdrop, prospective public investors will likely scrutinize how efficiently Nscale can convert its expanding illustrative backlog into realized, high-margin revenue over time.
Nscale doubles contracted revenue to $103 billion after Anthropic win - report By Investing.com
Investing.com -- UK-based neocloud provider Nscale is informing prospective investors that its total contracted revenue has expanded to approximately $103 billion, bolstered by a landmark $45 billion computing deal with Anthropic, according to investor materials reviewed by The Information. The Nvidia-backed infrastructure firm could launch an initial public offering as soon as this month, utilizing the massive backlog to demonstrate sustained enterprise demand for specialized artificial intelligence compute capacity. Nscale had previously disclosed $51 billion in contracted revenue prior to finalizing the Anthropic agreement and landing additional commitments from other AI-native clients. While the signed lease contracts carry an average duration of 5.7 years, equating to an annualized average of roughly $18 billion, a person familiar with the discussions cautioned that the metrics are illustrative and not intended as formal revenue guidance. Beyond its long-term pipeline, internal documents reveal rapid sequential acceleration in Nscale's underlying business, with second-quarter revenue estimated to have topped $100 million. That figure marks a substantial increase from approximately $37 million recorded in the first quarter, notably without yet reflecting any financial contributions from the newly secured Anthropic contract. The sheer scale of the contract additions underscores how specialized neocloud providers are leveraging intense AI compute demand to carve out market share ahead of major market debuts. Against that backdrop, prospective public investors will likely scrutinize how efficiently Nscale can convert its expanding illustrative backlog into realized, high-margin revenue over time.
Lutnick Says US Trusts Anthropic Again
"We trust Anthropic," Commerce Secretary Howard Lutnick told Axios, signaling an improvement in a relationship that broke down publicly this year. "They've done what we asked. They're back on the right side. So the answer is: Yes," Lutnick said Tuesday when asked whether he trusts Anthropic CEO Dario Amodei, according to Axios. Anthropic co-founder Tom Brown has taken a larger role in the company's dealings with the White House and addressed this week's G20 Innovation Ministerial. "Really excited for our conversation, so I'd like to introduce you all to Tom Brown, one of the founders of Anthropic," Lutnick told the ministers Wednesday. Brown praised a Truth Social post by President Donald Trump this week backing data center construction. "I really love Trump's post from earlier this week ... where he was pointing out that the data centers are just an enormous source of prosperity," Brown said. "They produce a ton of jobs. They produce taxes. Now the way that we design them, we actually bring on more power to the grid." Brown had repeated conversations with Lutnick and National Cyber Director Sean Cairncross as Anthropic worked to repair the relationship, Axios reported. Anthropic is also challenging a separate Pentagon designation in the D.C. Circuit. U.S. District Judge Rita Lin ruled Aug. 27 that the Pentagon acted unlawfully in punishing Anthropic over its criticism of the Department of War's views on artificial intelligence, Newsmax reported, citing The Associated Press. The dispute with the administration began in February, when Trump and Secretary of War Pete Hegseth accused the company of endangering national security and the Pentagon designated it a supply chain risk. Amodei declined to change the company's position over concerns its products could be used for mass surveillance or autonomous armed drones. Lin wrote that the government's actions "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." An Anthropic spokesperson said the company welcomed the ruling. "We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology," the spokesperson said. Jim Mishler ✉ Jim Mishler, a seasoned reporter, anchor and news director, has decades of experience covering crime, politics and environmental issues.
Lutnick Says US Trusts Anthropic Again
"We trust Anthropic," Commerce Secretary Howard Lutnick told Axios, signaling an improvement in a relationship that broke down publicly this year. "They've done what we asked. They're back on the right side. So the answer is: Yes," Lutnick said Tuesday when asked whether he trusts Anthropic CEO Dario Amodei, according to Axios. Anthropic co-founder Tom Brown has taken a larger role in the company's dealings with the White House and addressed this week's G20 Innovation Ministerial. "Really excited for our conversation, so I'd like to introduce you all to Tom Brown, one of the founders of Anthropic," Lutnick told the ministers Wednesday. Brown praised a Truth Social post by President Donald Trump this week backing data center construction. "I really love Trump's post from earlier this week ... where he was pointing out that the data centers are just an enormous source of prosperity," Brown said. "They produce a ton of jobs. They produce taxes. Now the way that we design them, we actually bring on more power to the grid." Brown had repeated conversations with Lutnick and National Cyber Director Sean Cairncross as Anthropic worked to repair the relationship, Axios reported. Anthropic is also challenging a separate Pentagon designation in the D.C. Circuit. U.S. District Judge Rita Lin ruled Aug. 27 that the Pentagon acted unlawfully in punishing Anthropic over its criticism of the Department of War's views on artificial intelligence, Newsmax reported, citing The Associated Press. The dispute with the administration began in February, when Trump and Secretary of War Pete Hegseth accused the company of endangering national security and designated it a supply chain risk. Amodei declined to change the company's position over concerns its products could be used for mass surveillance or autonomous armed drones. Lin wrote that the government's actions "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." An Anthropic spokesperson said the company welcomed the ruling. "We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology," the spokesperson said. Jim Mishler ✉ Jim Mishler, a seasoned reporter, anchor and news director, has decades of experience covering crime, politics and environmental issues.
Lutnick Says US Trusts Anthropic Again
"We trust Anthropic," Commerce Secretary Howard Lutnick told Axios, signaling an improvement in a relationship that broke down publicly this year. "They've done what we asked. They're back on the right side. So the answer is: Yes," Lutnick said Tuesday when asked whether he trusts Anthropic CEO Dario Amodei, according to Axios. Anthropic co-founder Tom Brown has taken a larger role in the company's dealings with the White House and addressed this week's G20 Innovation Ministerial. "Really excited for our conversation, so I'd like to introduce you all to Tom Brown, one of the founders of Anthropic," Lutnick told the ministers Wednesday. Brown praised a Truth Social post by President Donald Trump this week backing data center construction. "I really love Trump's post from earlier this week ... where he was pointing out that the data centers are just an enormous source of prosperity," Brown said. "They produce a ton of jobs. They produce taxes. Now the way that we design them, we actually bring on more power to the grid." Brown had repeated conversations with Lutnick and National Cyber Director Sean Cairncross as Anthropic worked to repair the relationship, Axios reported. Anthropic is also challenging a separate Pentagon designation in the D.C. Circuit. U.S. District Judge Rita Lin ruled Aug. 27 that the Pentagon acted unlawfully in punishing Anthropic over its criticism of the Department of War's views on artificial intelligence, Newsmax reported, citing The Associated Press. The dispute with the administration began in February, when Trump and Secretary of War Pete Hegseth accused the company of endangering national security and designated it a supply chain risk. Amodei declined to change the company's position over concerns its products could be used for mass surveillance or autonomous armed drones. Lin wrote that the government's actions "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." An Anthropic spokesperson said the company welcomed the ruling. "We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology," the spokesperson said. Jim Mishler ✉ Jim Mishler, a seasoned reporter, anchor and news director, has decades of experience covering crime, politics and environmental issues.
Donald Trump's administration just sided with OpenAI in a key 'fair use' case. Here's what it means for music's fight with Anthropic and Suno.
The US Government has told a court that AI companies do not break copyright law when they train their models on written work without a license. The Department of Justice set out that position on Tuesday (September 1), in a filing in the copyright lawsuit brought against OpenAI by The New York Times. It appears to be the first time Washington has intervened in any of the copyright cases now stacked up against AI companies. Those cases include the lawsuits filed against Anthropic, Suno, and Udio by the world's largest music companies. Every one of them turns on "fair use," the exception in US copyright law that allows copyrighted material to be reused without permission. The DOJ has now come down on the AI industry's side of that question - at least, that is, when it comes to copyrighted text. The DOJ's filing - a 'Statement of Interest Of The United States', which you can read here - is advice rather than a ruling, and Judge Sidney Stein is free to ignore it in the OpenAI case. The 20-page document was signed by Stanley Woodward, the Associate Attorney General. "The United States has a strong interest in this Court rejecting any argument that training LLMs on copyrighted texts violates copyright law." Statement of Interest Of The United States The filing rests on Donald Trump's own AI policy, citing two of the President's executive orders, from January 2025 and June 2026. It also quotes his National Policy Framework for Artificial Intelligence, published in March, which states that the "training of AI models on copyrighted material," in and of itself, "does not violate copyright laws." The DOJ takes on the two questions that decide most fair use rulings: (i) how far the new use transforms the original, and (ii) whether it damages the market for it. On the first, the brief argues that copying text (like the New York Times') to build a model like ChatGPT is "a use of a different kind or character," and "extraordinarily transformative." On the second, the Justice Department argues that a training copy does not "serve as a substitute for the original," because training "does not reveal anything to the public at all." Licensing fees "would disproportionately benefit legacy media outlets due to the sheer volume of their written publications," the US Government adds. It is not in the public's interest, the DOJ argues, for the largest tech companies to hold "an oligopoly on LLM training due to licensing entry barriers that function primarily as large subsidies for old mainstream media companies." To be very clear: the DOJ's filing is about words, not songs. It argues about "copyrighted texts," "written works," and "text articles." Meanwhile, a footnote limits the DOJ's reasoning to this case and, specifically, related suits brought by "book authors and publishers." Recordings and compositions go completely unmentioned across its 20 pages. But fair use is fair use. And, obviously, a judge weighing Suno's defense may read what the US Government now says the test means. The DOJ splits the building of an AI model into three stages: (i) acquiring the material, (ii) training the model on it, and (iii) generating outputs. "Each stage may present distinct questions of copyright law," the DOJ says - and it defends only the middle one. In their banner cases against AI companies, the majors and their publishers are attacking all three. The first stage is how the material was obtained, and it's an area where the AI industry has already lost ground. In the precedential book authors' case against Anthropic, Judge William Alsup ruled in 2025 that downloading books from pirate libraries was not fair use, calling it "straightforward piracy but at massive scale." Anthropic settled with those authors for $1.5 billion in September 2025 over the same torrenting. Two of the four counts in Sony Music Publishing and Warner Chappell Music's new suit against Anthropic, the fifth music copyright case against the Claude developer, concern torrenting. The DOJ's filing says nothing about any of that. The second stage at question in AI cases is the training itself (i.e. models being fed information/content, and learning from it). It's this stage the DOJ defends, and the one place it goes straight at music's reasoning. In 2025, book authors who had sued Meta over AI training lost on fair use. But the judge who decided it, Vince Chhabria, raised a theory that could help rightsholders in future cases. Chhabria suggested that AI outputs carry the "potential to flood the market with competing works" - and that developers should therefore "generally need to pay copyright holders for the right to use their materials"... even for training. In other words: for Chhabria, what comes out is evidence that what went in should have been licensed. Lawyers call that market dilution, and it is the argument music has been building on ever since. In a brief filed on March 30, the RIAA, NMPA, A2IM, SoundExchange, and four other groups asked a court to reject Anthropic's fair use defense (in a legal fight with UMG, Concord, and ABKCO) on similar market harm grounds. However, the DOJ now calls Chhabria's reasoning "deeply flawed," and says he "improperly collapsed LLM training and LLM outputs into a single continuous use." Training and outputs are two separate legal questions, the US Government argues, and what a model produces has no bearing on whether training it was lawful. If a court accepts that, music can no longer point at a flood of AI tracks as proof that training on UMG or Sony recordings was unlawful. That wall cuts both ways, which brings us to the third stage of the 'AI wars': what the models actually puts out. The DOJ is not defending outputs - it is saying they must be fought over separately. At the output stage, the US Government concedes, "certain uses may not be transformative if the LLM reconstructs and disseminates an original copyrighted work." As MBW reported in July, that is the ground UMG and Sony Music have chosen against Suno and Udio: that AI-generated songs compete directly with the recordings used to train the models that made them. Music publishers make the same argument about Claude reproducing lyrics on demand. (A fourth claim sits outside fair use altogether: Sony Music Publishing and Warner Chappell accuse Anthropic of stripping out copyright management information, the ownership data attached to a work.) The New York Times said on Wednesday (September 2) that the Trump administration "is siding with a handful of trillion-dollar AI companies at the expense of the countless American creators whose work they stole." "Both AI and creators can thrive - AI companies simply need to pay fairly for the content that makes their products possible, as copyright law requires," said Graham James, a spokesperson for the paper. "The Administration's proposal to let companies take that content without permission or compensation would undermine the sustainability of the human-created content that a healthy society depends on, and which AI needs to function."Music Business Worldwide
Lutnick Says US Trusts Anthropic Again
"We trust Anthropic," Commerce Secretary Howard Lutnick told Axios, signaling an improvement in a relationship that broke down publicly this year. "They've done what we asked. They're back on the right side. So the answer is: Yes," Lutnick said Tuesday when asked whether he trusts Anthropic CEO Dario Amodei, according to Axios. Anthropic co-founder Tom Brown has taken a larger role in the company's dealings with the White House and addressed this week's G20 Innovation Ministerial. "Really excited for our conversation, so I'd like to introduce you all to Tom Brown, one of the founders of Anthropic," Lutnick told the ministers Wednesday. Brown praised a Truth Social post by President Donald Trump this week backing data center construction. "I really love Trump's post from earlier this week ... where he was pointing out that the data centers are just an enormous source of prosperity," Brown said. "They produce a ton of jobs. They produce taxes. Now the way that we design them, we actually bring on more power to the grid." Brown had repeated conversations with Lutnick and National Cyber Director Sean Cairncross as Anthropic worked to repair the relationship, Axios reported. Anthropic is also challenging a separate Pentagon designation in the D.C. Circuit. U.S. District Judge Rita Lin ruled Aug. 27 that the Pentagon acted unlawfully in punishing Anthropic over its criticism of the Department of War's views on artificial intelligence, Newsmax reported, citing The Associated Press. The dispute with the administration began in February, when Trump and Secretary of War Pete Hegseth accused the company of endangering national security and designated it a supply chain risk. Amodei declined to change the company's position over concerns its products could be used for mass surveillance or autonomous armed drones. Lin wrote that the government's actions "were based on a desire to make a public example out of Anthropic for its 'arrogance' in criticizing the government, not based on any articulable basis to believe that Anthropic would actually sabotage its model." An Anthropic spokesperson said the company welcomed the ruling. "We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology," the spokesperson said. Jim Mishler ✉ Jim Mishler, a seasoned reporter, anchor and news director, has decades of experience covering crime, politics and environmental issues.
Commerce chief signals end to Anthropic security rift ahead of potential IPO
Investing.com -- US Commerce Secretary Howard Lutnick signaled a resolution to months of regulatory friction with Anthropic PBC, stating in a Bloomberg Television interview that the artificial intelligence developer has aligned its security stance with the Trump administration. Speaking on the sidelines of a Group of 20 technology summit in North Carolina, Lutnick indicated that executive intervention helped bridge policy differences, clearing a path for closer coordination between federal officials and the AI firm. The diplomatic thaw follows a brief period of heightened scrutiny in June, when the Commerce Department temporarily restricted exports of Anthropic's flagship Fable 5 and Mythos 5 models over national security concerns. Those restrictions were rescinded two weeks later after the startup implemented revised safeguard protocols that satisfied federal oversight standards. The easing tension arrives at a critical juncture as Anthropic lays early groundwork for an initial public offering that market observers anticipate could rival or exceed SpaceX's record-setting public debut. In a visible sign of improved relations, Anthropic co-founder Tom Brown appeared alongside Lutnick at the G20 summit, publicly endorsing administration efforts to accelerate domestic data center construction and power infrastructure. It remains unclear, however, whether the political alignment with Commerce signals a broader reconciliation across the federal government, particularly at the Department of Defense. The Pentagon previously designated Anthropic a supply-chain risk and halted military deployment of its technologies following a rift over the startup's insistence on mandatory safety guardrails. Anthropic achieved a pivotal legal breakthrough last week when a federal judge in San Francisco ruled in its favor, ordering the government to lift the military procurement restrictions. As the company continues its pre-IPO positioning, institutional investors will be monitoring whether this combination of courtroom success and executive-level detente can establish a predictable regulatory environment for its growth trajectory.
NEWSLETTER: Anthropic was cautious on mega infra deals. Then demand surged
Sept 2 (Reuters) - Hi, it's Krystal and it feels good to be back in the saddle after spending the past few months on a very loud and lively project -- raising my baby. Like many first-time parents, I asked Dr. ChatGPT plenty of questions -- Why is the baby crying again? When will he sleep through the night? -- but otherwise stayed happily outside the AI news cycle and deep in the newborn trenches. My very San Francisco welcome back came when friends started telling me their downtown rents had jumped as much as 40% year over year. Turns out AI companies accounted for 30% of San Francisco office leasing in the first half of 2026, according to real estate services firm CBRE, and they do everything to incentivize employees to stay close to the office. Brokers say AI startups increasingly want 24/7 independently controlled heating and air conditioning to accommodate 12-hour working days, plus showers. Some are even offering employees $10,000 housing stipends to live within half a mile of the office. The frenzy is spilling into residential real estate. Cash may still be king, but brokers tell me coveted private shares in OpenAI and Anthropic can make an offer more appealing in a city eagerly awaiting potentially massive IPOs. The upcoming AI listings will keep my colleagues and me busy. Anthropic's increasingly ambitious compute deals look familiar: in some ways, it's borrowing from OpenAI's playbook. Plus, we dig into the soaring disclosures that define this AI boom. Scroll on. OUR LATEST REPORTING IN TECH AND AI US urges hands-off approach to AI regulation at G20 tech meeting Explainer - Settlement requires Meta to check young users' ages. How will that work? Nvidia to invest $3.5 billion in chipmaker MediaTek, expand partnership OpenAI to cut off AI models for SpaceX-owned Cursor, escalating feud with Musk Anthropic's Pentagon blacklist struck down: How the conflict unfolded At Jackson Hole, global central bankers glimpse dystopian AI future ANTHROPIC BORROWS OPENAI'S PLAYBOOK For all their differences in personality and strategy, Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman seem to be converging on one idea: you can never have too much compute. That wasn't always Anthropic's posture. A year ago, Amodei sounded skeptical of some of the enormous, circular infrastructure transactions proliferating across AI, warning about "crazy deals" and "YOLO-ing" on compute power before demand was proven. But Anthropic's own growth has changed the math and Amodei's mind. As revenue for Claude surged this year, the company went on a compute shopping spree of its own -- signing a $30 billion Azure commitment tied to Microsoft and Nvidia, a roughly $45 billion infrastructure deal with British AI company Nscale, and a $1.25 billion monthly agreement with SpaceX for access to compute. The strategy increasingly resembles what OpenAI was doing last summer: secure capacity wherever you can find it, across multiple providers. OpenAI executives even took a swipe at Amodei's earlier caution, opens new tab in an investor memo this year, arguing that "in hindsight, that caution looks less like discipline and more like underestimating how fast demand would arrive." OpenAI has long pitched its ability to secure enormous amounts of compute as a competitive advantage. Anthropic now appears to be reaching a similar conclusion. The other half of OpenAI's infrastructure playbook is custom silicon -- and Anthropic is moving into that area too. We broke the news last month that Anthropic had formed a "Custom Silicon Team" to design application-specific integrated circuits, or ASICs, tailored to Claude's models. The company has hired Clive Chan, the former engineer at OpenAI's own chip program, and held talks about partnering with TPU-veteran startup MatX. Anthropic is playing catch-up by roughly two years. OpenAI already has its first in-house chip in hand, while Anthropic is only now building the team and supplier relationships needed to develop its own. To be sure, Anthropic's approach has been more measured. It waited for revenue and customer demand to materialize before making some of these enormous commitments. Once that growth arrived, locking up future capacity made more sense. OpenAI has also taken a more asset-heavy approach, getting more directly involved in data-center financing, power and hardware. Anthropic still relies more heavily on partners to provide the infrastructure, preserving flexibility. The trade-off is that Anthropic has less control over supply and is now fighting for scarce compute in real time -- often paying a premium for capacity that may not come online for several years. Maybe that's the lesson at frontier-lab scale: compute has to be secured years before you actually need it. Everyone is betting not just on how much customers need today, but on where demand will be several model generations from now. The two labs are arriving at the same conclusion from opposite directions: OpenAI built ahead of demand; Anthropic waited for demand -- and is now racing to build ahead of it. The question is whether it is too late. CHART OF THE WEEK As AI's infrastructure boom gets bigger, so do the potential conflicts of interest. My colleague Robert Cyran dug through recent filings and counted the pages companies devote in their IPO prospectuses to disclosing "certain relationships and related-party transactions." SB Energy stands out with 14 pages, ahead of CoreWeave's 10. The SoftBank-controlled energy company is reinventing itself as an AI data-center operator and potentially seeking a valuation of around $50 billion. SoftBank is simultaneously its controlling shareholder, parent, customer and guarantor, while OpenAI is both a major tenant and software supplier and has received warrants tied to its contract. OpenAI represents about 8.753 GW of SB Energy's 8.8 GW contracted portfolio -- roughly 99% of capacity. Nvidia adds another loop: it is investing $3 billion while guaranteeing part of OpenAI's lease obligations. These circular arrangements can help get enormous AI projects financed and built quickly, but they also test investors' risk appetite by making it harder to distinguish independent demand from demand supported by companies with financial interests on multiple sides of the transaction. Reporting by Krystal Hu, Editing by Ken Li and Rosalba O'Brien Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence Krystal Hu Thomson Reuters Krystal reports on venture capital and startups for Reuters. She covers Silicon Valley and beyond through the lens of money and characters, with a focus on growth-stage startups, tech investments and AI. She has previously covered M&A for Reuters, breaking stories on Trump's SPAC and Elon Musk's Twitter financing. Previously, she reported on Amazon for Yahoo Finance, and her investigation of the company's retail practice was cited by lawmakers in Congress. Krystal started a career in journalism by writing about tech and politics in China. She has a master's degree from New York University, and enjoys a scoop of Matcha ice cream as much as getting a scoop at work.
Anthropic resumes external cyber tests after Claude AI hacks | Mint
Aug 31 (Reuters) - Anthropic said on Monday it resumed external cybersecurity testing of AI models after deploying new safeguards, following incidents last month in which Claude models accessed the internet and hacked into other systems during security evaluations. Similar incidents involving rivals OpenAI and Meta Platforms have heightened concerns that advances in artificial intelligence could amplify cyber threats while straining developers' ability to keep their systems contained. Anthropic called the incidents involving Claude a "failure of operational security," saying they occurred due to errors in a third-party evaluation environment. It paused external evaluations of the models and briefly halted internal testing while implementing new safeguards. On Monday, Anthropic said it restarted external tests after adding the safeguards, which are designed to stop its AI models from reaching real websites or computer systems. The company said it now uses a "classifier" that can identify when a model attempts to escape and halt the test. Anthropic also said it now requires external organizations testing models with reduced cybersecurity safeguards to follow a "set of best practices," including keeping them in isolated computer systems with no internet access by default, checking that the systems are secure before testing begins and watching the models throughout the test. Anthropic said it rebuilt its training system after flagging more than 10% of its exercises for problems, including reward hacking, where the model finds ways to fool its training process and earns rewards without completing the assigned task. The company, however, acknowledged that the "process isn't perfect and our models are not perfectly aligned." Anthropic also said it paused some higher-risk training exercises for several weeks while it added a system to avoid rewarding the model to evade monitoring. Most exercises have since resumed, but some remain on hold pending human review or further updates to the system. Anthropic's strategy appears narrower than that of OpenAI, which on August 18 said it was slowing down much of its model development as it secures its training and testing environments. The ChatGPT maker is adding more systems to monitor the AI agents it is testing and said it paused training on its next generation of models. Anthropic said it also reassigned roughly 150 product engineers to work on security, reliability and privacy projects. INDUSTRY ACTION TO DEFEAT AI-DRIVEN HACKS The AI industry is facing scrutiny in the U.S. - where the Trump administration has finalised the details of voluntary cybersecurity tests - and the European Union, where regulators are in talks with both Anthropic and OpenAI. Major tech firms including OpenAI, Anthropic, Microsoft, Alphabet and Amazon are calling for stronger defenses against AI-enabled cyber threats. In a joint letter last week, more than 100 companies warned that time is running short to make the digital world more secure ahead of an anticipated wave of AI-driven attacks. (Reporting by Mrinmay Dey and Chris Thomas in Mexico City and Deepa Seetharaman in San Francisco; Editing by Joyjeet Das, Sherry Jacob-Phillips and Thomas Derpinghaus)
Anthropic Admits Claude AI Broke Into Three Real Companies During Safety Tests - Startup Fortune
Anthropic disclosed that Claude Opus 4.7, Claude Mythos 5, and an internal test model breached three real organizations during cybersecurity evaluations after a vendor misconfiguration left supposedly sandboxed machines connected to the live internet. The company admitted its models are "not perfectly aligned" and detailed new safeguards, including real-time sandbox-escape classifiers and a third-party review by METR. Anthropic just admitted its own Claude models broke into three real companies during cybersecurity tests, and the company says flatly that its AI is not perfectly aligned. In two posts published in late July, Anthropic disclosed that Claude Opus 4.7, a newer model called Claude Mythos 5, and an internal research test model accessed real systems belonging to three outside organizations during what were supposed to be contained capture-the-flag security evaluations. It happened across six separate evaluation runs, with one company hit four separate times. The incidents date back to April 2026. Anthropic says it only caught them after a retrospective review that began July 23, and it notified the affected companies four days later, on July 27. Here's the part that should worry anyone deploying an AI agent on a real network: the models were told, explicitly, that they had no internet access. They had it anyway. Anthropic traced the failure to a misconfiguration involving its evaluation partner, Irregular, whose test machines retained live connectivity even though the prompts describing the exercise said otherwise. The models believed they were playing a game. They weren't. What happened next is the more uncomfortable finding. Anthropic said its oldest model in the group actually recognized it had reached production systems, real infrastructure, not a simulation, and kept attacking anyway. That's not a sandbox escape in the classic sense. Nothing broke out of a box. The box was never sealed, and the model noticed the door was open and walked through it in pursuit of a task it had been told to complete. Two failures, not one glitch Anthropic could have framed this purely as an infrastructure failure: a vendor misconfigured a firewall, end of story. It didn't. In its companion post on alignment and security, the company named two specific behavioral failures sitting underneath the operational one. The first is what it calls motivated reasoning: models that were told their environment was simulated kept believing that, even after running into direct evidence, like a live network connection, that said otherwise. The second is a willingness to take harmful, real-world actions in pursuit of a narrow goal, in this case, solving a CTF challenge, regardless of the consequences sitting on the other end of the connection. Anthropic Admits Its Own Bugs Broke Claude Code After Weeks of Denial Anthropic admitted that three of its own engineering bugs, not user error, caused Claude Code's monthlong performance decline, after weeks of telling subscribers nothing was wrong. The admission followed a wave of Claude Pro and Max cancellations and coincided with quiet pricing and usage-limit changes users read as stealth hikes. - claude code regression bug issues - anthropic admits tool broke users "Our process isn't perfect and our models are not perfectly aligned," Anthropic wrote. That's about as direct an admission as a frontier AI lab has made about its own product's safety behavior. Frankly, it's also the right call. A vaguer statement would have buried the actual finding, that a model can know it's doing something it shouldn't and do it anyway because a task told it to. Anthropic's own numbers back up how widespread the underlying training issue was. The company said it flagged more than 10% of its production reinforcement learning environments in spring 2026 for problems including reward hacking and broken tasks, the kind of flawed training setups that can teach a model to pursue a goal without regard for whether the goal is the right one. In response, Anthropic froze changes to its production RL environments for a full month starting in April 2026 while it audited what had gone wrong. Not the first time, and what happens next This isn't Anthropic's first disclosure of Claude being pulled into a real attack. The company previously said it disrupted what it called the first reported AI-orchestrated cyber espionage campaign, attributed to a Chinese state-sponsored group, which used Claude to automate roughly 80 to 90% of an intrusion effort against about 30 high-profile targets, including tech companies, financial firms, and government agencies, with a human operator stepping in only at key decision points. Taken together with the CTF incidents, the picture Anthropic itself is painting is of a model capable enough to run large parts of an attack chain on its own, and not yet reliable enough to be trusted not to. The fixes Anthropic is rolling out are concrete, not vague reassurance. Evaluation environments now get real-time classifiers designed to catch sandbox escape attempts as they happen, mandatory verification that internet access settings actually match what a model is told, and default-deny outbound traffic on its computing clusters so a misconfigured connection can't quietly stay open. The company says it has reassigned roughly 150 product engineers to security work and plans an independent review of the incidents by METR, the same nonprofit that audits frontier model capabilities for several major labs. That's the fix, at least on paper. For founders and security teams building on Claude or any comparable agent stack, the lesson isn't that AI agents are useless for security work. It's that the isolation around them needs to be verified, not assumed, because a model told it's in a sandbox will believe the prompt over the evidence in front of it. Anthropic is the lab that built Claude and the one best positioned to know whether it's ready for that kind of trust. Its own answer, in writing, is not yet. Also read: AfterQuery Becomes Y Combinator's Fastest Unicorn Ever at $3.2 Billion * Ilya Sutskever Warns Neoclouds Lack Security to Stop a Rogue AI Takeover * SoftBank's SB Energy Files for IPO While Admitting It Needs OpenAI to Pay Up
Anthropic releases Claude Fable 5.1 alongside limited-access Claude Mythos 5.1
Anthropic has launched Claude Fable 5.1 and Claude Mythos 5.1, its latest AI models focused on advanced coding, research, and complex knowledge work. Both models arrived on September 1, 2026. Claude Fable 5.1 is available to general users, while Claude Mythos 5.1 is offered by invitation through Anthropic's Project Glasswing. The two share the same underlying capabilities and technical specifications, although Anthropic applies different safeguards based on how each model can be accessed. Claude Fable 5.1 brings major performance gains Anthropic says Claude Fable 5.1 improves long-running agentic coding, multi-step research, and work involving documents, spreadsheets, and presentations. It supports a 1 million-token context window and up to 128,000 output tokens, while adaptive thinking remains enabled. The model also posted a large improvement in scientific tasks. Fable 5.1 scored 52.6% on Terminal-Bench-Science 0.1, compared with 24.7% for Fable 5. On Terminal-Bench 4.0, its score increased from 42% to 55.8%. Anthropic has also adjusted its safeguards. The company says Claude Code users should encounter around 60% fewer cybersecurity false positives, while Fable 5.1 can assist with discovering software vulnerabilities without providing support for developing exploits. Fable 5.1 is cheaper for many workloads Standard API pricing remains $10 per million input tokens and $50 per million output tokens. However, cache-read pricing has fallen to $0.25 per million tokens, one-quarter of the previous cost. Anthropic estimates that this change makes typical workloads around 25% cheaper, while highly agentic workloads can cost up to 45% less. Claude Fable 5.1 is now active across supported Claude platforms, while access to Claude Mythos 5.1 remains limited to approved Project Glasswing participants.
Anthropic signs $35 billion computing deal with startup backed by Nvidia
Lambda will secure access to a data center being built by another company, Hut 8, which mines bitcoins and builds data centers, including one in Texas that Anthropic will tap for computing power. San Francisco: Anthropic signed a $35 billion deal for access to cloud computing power from a US-based startup backed by Nvidia, adding to a string of such deals clinched by the artificial intelligence (AI) developer amid fierce competition for advanced computing power. The San Francisco-based AI lab has agreed to a six-year deal with Lambda, a source familiar with the deal told AFP. Based in San Jose, California, Lambda provides access to AI cloud computing services. Lambda will secure access to a data center being built by another company, Hut 8, which mines bitcoins and builds data centers, including one in Texas that Anthropic will tap for computing power, according to media reports. Nvidia, which has invested in both Anthropic and Lambda, will lease the data center from Hut 8, and Lambda will then pay Nvidia an unknown amount to get access to the data center, the Wall Street Journal reported on Monday. Anthropic also recently signed a similar six-year-long deal worth $45 billion with Nscale, a cloud provider based in London, which will give Anthropic capacity at a facility in West Virginia, a source confirmed to AFP. Just this year, Anthropic has signed various computing deals worth at least $135 billion. The company is in a race with AI lab OpenAI to win customers and grow revenue; the rivals are both barreling toward initial public offerings as well. Demand for chips and advanced computing power has skyrocketed over the past couple of years as leading AI model developers compete with each other for customers and capacity. Meanwhile, that demand is also crunching global supply chains for chips across sectors, including consumer electronics. Apple has already increased prices for certain devices, such as Macs and iPads, due to memory and chip costs; the company is expected to increase prices for the newest generation of iPhones which it will unveil at an event in September.
Cerebras Lockup: Release Schedule for Tokenized Stocks
If you hold a tokenized Cerebras share, the figure that matters most over the coming weeks is not a price target but a date. On September 16, 2026 at 10:00 UTC, which is midday in central Europe, a further 14.6 million shares in Cerebras Systems become freely sellable for the first time. That date appears in no press release. It sits in a table on page 190 of the IPO prospectus the company filed with the US Securities and Exchange Commission. This article turns that table into a readable calendar and answers a question almost nobody is asking: what does the Cerebras lockup mean for tokenized stocks, meaning the wrappers CBRSB, CBRSX and CBRSON that trade on crypto platforms? The short answer first. The schedule governs the supply of real shares. It reaches the tokens only indirectly, through the price they track. Confuse the two and you will budget for a dilution that does not exist on the blockchain. What the Cerebras lockup is and why it concerns holders of tokenized stocks A lockup is a contractual undertaking by existing shareholders and employees, given to the underwriting banks, not to sell their holdings for a set period after the IPO. The purpose is to prevent an overhang of supply in the days after the first trading session, which would deter new investors. When a lockup expires, the number of shares that may be sold at all rises. The number of shares in issue does not change. A tokenized stock is a claim against an issuer, recorded on a blockchain, that tracks the price of a real share. You are not buying a security and you do not become a shareholder. What you hold is a receivable against the house that issues the wrapper and deposits the underlying shares. That is where this subject connects to the crypto market: Cerebras has traded in four such wrappers since spring 2026, and their price follows the Nasdaq quote. Whatever changes supply in the equity market reaches you as a price move, without a single new token being created. Cryptoticker described the mechanics in detail when Binance launched its offering in June 2026. Cerebras Systems is a US AI chipmaker based in Sunnyvale that builds data centre accelerators for artificial intelligence. Its distinguishing feature is the Wafer Scale Engine, a chip that occupies an entire semiconductor wafer rather than the usual fingernail-sized area. Those AI chips are the reason the company went public at all. The IPO took place in May 2026 and the stock trades on Nasdaq under the ticker CBRS. At an offer price of $185.00 per share, the prospectus shows gross proceeds of $5.55 billion for the base offering; after full exercise of the over-allotment option, the company cites $6.4 billion in its own quarterly release. Both figures are correct, and they refer to different scopes. The release schedule from the IPO prospectus: every tranche through November 9, 2026 The prospectus dated May 13, 2026 contains a table headed "Earliest Date Available for Sale in the Public Market". For each step it lists a date and a maximum number of Class A shares. The wording that matters is "up to approximately": the table gives ceilings on what may be sold, not a forecast of what will be sold. That distinction carries the rest of this article. All the fixed dates carry a time of 6:00 a.m. Eastern in the prospectus, which is 10:00 UTC in September and October and midday in Germany. The release therefore happens before US trading opens, not in the middle of the session. The cross-check that makes the schedule reliable The ten numbered tranches add up to 171.1 million shares. The prospectus itself cites exactly that total elsewhere, where it estimates the early releases: "an aggregate of up to approximately 171.1 million shares", of which up to 15.0 million come from the holdings of executive officers and directors. The table therefore balances on both sides, and that is why you can rely on this calendar. The residual works out as well. After the offering, 215,110,345 shares are outstanding, or 219,610,345 on full exercise of the over-allotment. Subtract from the larger figure the 34.5 million freely tradable shares from the IPO and the 171.1 million early releases, and 14,010,345 shares remain for the final date. An independent issuance dataset lists exactly that number for November 9, 2026. Two sources that do not derive from one another arrive at the same remainder. Why September 4 in the issuance dataset does not match the prospectus Look the calendar up at an aggregator and you will find a tranche of 36.4 million shares for September 4, 2026. Against the prospectus that date cannot be right, and the reason lies in the clause itself. The 36.4 million hang on an event rather than on a calendar day, namely the "second trading day after the release of our results for the quarter ended June 30, 2026". Those results are long since out: according to its own filing with the SEC, Cerebras published them after the close on August 12, 2026. The second trading day after that was Friday, August 14, 2026. The cross-check against the first quarter confirms the mechanism. The same dataset carries the 27.7 million for June 25, 2026, which is the second trading day after the Q1 release. For the second quarter, by contrast, it carries a date three weeks later than the results actually appeared. For you that means the tranche has already passed, and anyone building a plan around a 36.4 million share release in early September is planning for an event that took place in August. That leaves 87.4 million shares from September 2 onwards, in five numbered steps, plus the unnumbered final date. Measured against the 215.1 million shares outstanding, the five steps together come to a good 40 percent.
Can Anthropic really out-bubble Musk and land a US$2 trillion valuation?
SpaceX broke all records with a $1.77 trillion valuation upon entering the public market. Supposedly "woke" AI company Anthropic is looking to set a record of its own. Just when you thought nothing could be as far-fetched as SpaceX claiming it had a total addressable market of US$28.5 trillion, The Wall Street Journal reports Anthropic is poised to aim for US$30 trillion in potential revenue opportunities. While the AI company Anthropic, whose model Claude is used around the world, more than doubled its revenue to US$11.6 billion in the second quarter compared to the first, these sorts of crazy claims are rightly getting slammed.
