Cryptopolitan

Alex Thorn, the head of firmwide research at Galaxy Research, shared Galaxy’s analysis on X, adding his voice to the debate regarding Strategy’s newly announced Digital Credit Capital Framework.
The new rules have sparked a debate about whether or not they will solve the company’s capital-structure problems or simply delay them.
How does Strategy’s new capital framework operate?
Strategy (Nasdaq: MSTR) recently disclosed a new “Digital Credit Capital Framework” in a 8-K regulatory filing.
Cryptopolitan reported that the framework grants the company formal permission to sell up to $1.25 billion worth of Bitcoin. Notably, the firm is facing a massive unrealized loss of roughly $14 billion on its holdings of 847,363 BTC.
The framework creates a formal USD reserve policy that introduces revised dividend terms for its STRC preferred shares, and authorizes separate repurchase programs for both preferred stock and MSTR common shares at $1 billion each.
The board has kept aside the company’s $2.55 billion cash reserve, restricting its use to preferred dividends and debt interest.
If the current spending rates of roughly $1.76 billion annually are retained, this reserve is expected to last for about 17 months. If the full authorized sale of Bitcoin were to be executed, total liquidity would stretch to approximately $3.8 billion, an amount that would cover about 26 months of obligations.
Alex Thorn of Galaxy Research pointed out that the core of the debate is whether these new rules actually solve Strategy’s capital-structure issues or simply delay them.
The company sold 32 BTC for about $2.5 million in its first-ever Bitcoin sale in late May to cover a dividend payment. JPMorgan recommends that Strategy sell its shares to raise money rather than sell Bitcoin.
Are investors buying MSTR shares?
MSTR shares climbed 12.6% to $92.68 on the Monday after the filing was disclosed, and then by Wednesday, the value had soared past $100. This figure represents a 27% increase from the prior Friday’s close. The STRC preferred shares also increased, closing at $87.87 on July 3.
Benchmark Equity Research sees the framework as a good thing. The firm kept its Buy rating on MSTR and set a price target of $570.
Strategy’s leaders, including Chairman Michael Saylor, say the overhaul is needed to strengthen the company’s credit. Saylor said that “digital credit requires liquidity, discipline, and active capital management.”
Strive, another firm pursuing a Bitcoin-backed capital structure, told investors on July 2 that they should not assume the company will automatically issue new shares of its SATA preferred stock at $100 par value, citing abnormal market conditions.
Strive’s chief risk officer Jeff Walton shared figures showing that short interest in SATA rose by about 1 million shares over the 30 days to June 30, with borrow costs spiking from 6.1% to 68.6% APR over the same period.
The smartest crypto minds already read our newsletter. Want in? Join them.

A 33-year-old CFO in Ahmedabad lost $81,700 or ₹78.99 lakh to a crypto investment fraud. It started as a match on a matrimonial app, police said. Investigators link the case to an organized network running the same script on victims across India.
A matrimonial match turns into a crypto trap
In March 2025, Hardik Chandrakantbhai Soni from Vadaj downloaded the Sangam app to find a partner. He matched with a profile named Harshitha Gondawith on May 8, 2025. Within days, the talk moved from the app to WhatsApp.
Police say she steered the chat toward cryptocurrency over the next few weeks. She told Soni about big trading wins. The winnings bought a house in central Berlin, and video proof to sell the lie. Soni had turned the pitch down at first. At last, she wore him down with her persistence.
She directed Soni to a trading site, m[.]bitcoin-on[.]com. To register, he had to give his name, email, phone number and driving license details. Then the customer support agent told him to send money to a rotating set of bank accounts. Every transfer receipt he took a screenshot of had a corresponding USDT credit on his dashboard.
Deposits were accepted from May 8, 2025 to January 26, 2026. The size of individual installments ranged from $104 or ₹10,000 to over $10,355 or ₹10 lakh.
The total was $81,700, or ₹78.99 lakh. The profits and balance shown on screen steadily increased throughout, but it was all a sham.
Soni wanted to take out his principal along with the claimed profits. Then the operators started demanding more money, calling the new charges taxes, processing fees and verification costs. Some of those extra payments cleared, but none of it was coming back to him.
Soni first registered a complaint on the National Cyber Crime Reporting Portal. An FIR was then registered with Ahmedabad Cyber Crime Police.
The complaint states Harshitha-FS64217032 is the matrimonial profile that manages the WhatsApp number and the full scam operation. Police do not believe there was only one victim in this case.
Other Indian professionals fall for crypto scams
Cryptopolitan also reported a similar case. Civil contractor Ramesh lost about $200,000, or ₹1.67 crore, to a woman who called herself Priyanka. Priyanka had met him on a matrimonial website, claiming she worked for a crypto trading firm based in Singapore.
A first deposit of ₹50,000, about $600, brought an instant ₹8,300 profit, about $100. That led him to much bigger transfers, both through UPI and bank channels. The second he tried to take money out, his wallet locked. The scammer then asked for another ₹25 lakh, about $30,000, and then disappeared.
Cryptopolitan reported another incident. Ashok Vijayvargiya, a chartered accountant who is 70 years old and also the Chief Returning Officer of the Madhya Pradesh Chamber of Commerce, was ripped off of ₹21.06 crore, which is about $2.2 million. He became friends with a con artist on social media who said her name was “Divya.” The con artist then put him through the same fake profit, frozen withdrawal scam.
Cybercrime expert and former IPS officer Prof. Triveni Singh says that criminals like matrimonial and social platforms because there is already a level of trust there before money comes into play. He says that before sending money to any investment platform, you should do your own research. He said it doesn’t matter how many property videos or screenshots are used to back up the return promises. They are red flags.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

The Massachusetts Senate voted Thursday to prohibit crypto ATMs. It was attached to an omnibus economic development bill.
The action comes as FBI figures show kiosk-scam losses statewide last year totaled ~$7 million. Consumer advocates and elderly residents say the unregulated machines have become a cash pipeline for scammers.
Crypto ATM scams walk victims to the kiosk
The machines look just like normal cash dispensers. Hundreds sit inside convenience stores, pharmacies and liquor stores across the state.
Norfolk County Sheriff Patrick McDermott said it usually starts with a call or text, and then the victim is pressured into moving savings into Bitcoin. Cash goes into the machine, a QR code from the caller gets scanned, and the deposit converts and routes to an anonymous wallet in seconds. Once the cash is in there, it’s gone for good.
The FBI says it got 296 complaints about kiosk scams in Massachusetts in 2025, causing $6,834,561 in damage. That’s about $19,000 in reported losses per day spread out. That brought the bureau’s national total to $389 million stolen in that way over the past year.
AARP Massachusetts led the charge for the crackdown. “A ban on crypto ATMs is now the most effective way to stop the ongoing damage,” state director Jen Benson said in a statement Thursday.
Massachusetts leaves crypto ATMs unregulated
Massachusetts does not have written rules for crypto kiosks. This prompted sheriffs and advocacy groups to push lawmakers to at least temporarily shut them down. Massachusetts is the only New England state that leaves these machines unregulated.
Vermont, Minnesota, Indiana and Tennessee banned the machines outright, and 28 other states restrict their use in some way.
The absence of local guardrails has lured operators in, says McDermott. He added that pulling the machines is the only lever available until real regulation exists.
But the vote on Thursday doesn’t change anything. The ban was added to the Senate’s broad economic development bill as an amendment. The bill was approved by the chamber late Thursday night.
A comparable language has already been looked at by the House twice, once in its version of the bill and once in an earlier budget proposal. It fell both times. The ban’s fate depends on how well the two houses can work together to pass their bills.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

Morgan Stanley’s Bitcoin exchange-traded fund (ETF) has managed to rake in almost $400 million since it was launched in April, which reflects the growing acceptance of cryptocurrencies among financial institutions. As regulations become clearer in the US and the Asian regions, traditional financial institutions are now ready to move on with their businesses in the digital asset economy.
For a wide range of institutional and wealth management clients, a Bitcoin product offered by banks represents what the cryptocurrency industry has always been missing—credibility based on regulations. The change symbolizes not just an increase in demand for investments, but also shows that clearer rules are diminishing the boundaries between conventional finance and cryptocurrencies.
Morgan Stanley’s fund becomes the first by a bank to near $400M
According to reports, the Bitcoin ETF from Morgan Stanley, which is being traded on the NYSE Arca, already has more than $391 million in assets under management. This ETF was introduced in April with over $33 million in assets as its opening amount, plus it holds the record of being the first Bitcoin ETF by a large bank. For this week alone, it raised $15.7 million, according to Farside Investors, and Bloomberg Intelligence’s senior ETF analyst described it as one of the greatest ETF launches of the year.
The inflows are notable since they took place at a tough time for the overall market. The US spot Bitcoin ETF has gathered a total of $274 million in inflows in the last week, but in the last few days Morgan Stanley’s firm has been the only asset manager to escape making net withdrawals.
Over the years, Morgan Stanley has constantly been developing its crypto presence. The first time it gave wealthy customers an option to invest in Bitcoin was in 2021, when it introduced Galaxy Digital funds. Its chairman, Ted Pick, stated that the bank holds talks with state authorities about how to expand its crypto services safely. In addition to this, in April, Amy Oldenburg, the head of digital assets at Morgan Stanley, argued that education is one of the biggest challenges facing the industry today, not product creation.
Washington still hasn’t finished the rulebook it promised
However, regulatory uncertainty still influences the US market.
The CLARITY Act has been passed by the House of Representatives but is still stuck in the Senate, where at least 60 votes are needed, including the support of some of the Democrats. Traders on Polymarket reduced the probability of the law getting passed in 2026 to 37% on July 22 after being in conflict over whether the enforcement of the proposed ethics provisions would be in the hands of the Justice Department or state attorneys general. Senator Angela Alsobrooks called the proposal from the White House “unserious.”
According to an analyst, if implemented, the legislation will empower the Commodity Futures Trading Commission to supervise the crypto spot markets, define “mature blockchain,” and determine whether digital tokens are regulated by the SEC or the CFTC. President Donald Trump has insisted that the Senate pass the legislation before its scheduled recess from August 11 onwards, but that may become more challenging than it seems.
Why banks watch the stablecoin yield fight so closely
A major concern is regarding the clause on stablecoins. Banks worry that interest payments on stablecoins will drive customers away from traditional lenders. Patrick Witt, a White House adviser for digital assets, has called it a possibility of “deposit flight,” urging a more controlled method of dealing with the issue of “idle yield” so the whole law, in general, is not compromised. This draft makes it illegal to pay interest on spare stablecoins but permits gaining rewards based on specific transactions.
After withdrawing support for an earlier Senate draft in January, Coinbase CEO Brian Armstrong endorsed the revised CLARITY Act in April, saying it was “time to pass” the legislation after months of bipartisan negotiations.
Financial implications of this legislation are huge, as evidenced by the figures recorded by Coinbase, as it had revenue from stablecoins of $305 million in Q1 of 2026 with average USDCs amounting to $19 billion. Furthermore, banks, asset managers and market infrastructure companies expand into crypto by means of custody, tokenization, ETFs, payments, and settlement and not by sticking to one line of business.
Singapore hands Coinbase the cleaner rulebook
As US politicians are still discussing cryptocurrency legislation, Singapore has already provided the regulatory clarity that many companies are seeking.
Coinbase plans to increase its headcount in Singapore from about 150 to 200 by the end of 2026. The move comes about three years after the company received its complete Major Payment Institution license from Singapore’s Monetary Authority in October 2023. Its various products in Singapore include the stablecoin in Singapore dollars, XSGD, whose reserves are maintained by DBS Bank and Standard Chartered.
The company is reducing headcount elsewhere, making its hiring push in Singapore a clear signal that it sees long-term institutional growth in well-regulated markets.
Bitcoin was trading near $64,096 as the news emerged, according to Bitcoin Magazine, showing little movement over the week. CoinShares head of research James Butterfill remained cautious, saying the firm sees “no significant upside potential from here.”
If you're reading this, you’re already ahead. Stay there with our newsletter.

Samsung Electronics and Broadcom have entered into a memorandum of understanding (MOU) in order to strengthen their collaboration in the fields of memory, foundry manufacturing and advanced packaging. They expect to spend more than $200 billion as a result of their partnership till 2030. The agreement comes at a time when many AI chipmakers are trying to secure their supplies of high-bandwidth memory (HBM), one of the key challenges in the industry.
The agreement was made public at an AI summit held in San Francisco where South Korean President Lee Jae Myung and representatives of prominent AI firms took part. The South Korean presidential office regarded the Broadcom collaboration as one of the initiatives to enhance the local AI semiconductor industry.
While Samsung acknowledged the anticipated financial projections and scale of the partnership, the firm and the President’s office of South Korea have declined to specify how many billions they would supply each year, what the prices would be, when shipments would be scheduled, or how the amount would be split across memory, foundry, and advanced packaging. Due to the lack of any order or buying agreement, the estimated value of $200 billion would represent a mere long-term strategy, rather than already determined revenues.
Why Broadcom buying memory matters beyond the two companies
Broadcom does not manufacture memory chips, but it has become one of the world’s largest designers of custom AI accelerators. Samsung said the broader partnership will support Broadcom’s next-generation AI chips through advances in HBM, foundry manufacturing and advanced packaging. Broadcom already develops Google’s Tensor Processing Units (TPUs) and Meta’s MTIA accelerators, both of which require large amounts of HBM.
This is the reason why the agreement has the potential to create waves in the AI sector. HBM continues to be one of the most constrained components in terms of availability in AI hardware. The Samsung-Broadcom deal can secure an entire manufacturing capability that other chip developers may want to access as well.
It is already clear that demand is on the rise. According to Seoul Economic Daily, SK Group Chairman Chey Tae-won has said Broadcom has been continuously asking for what he calls an astonishing amount of memory chips, emphasizing the way hyperscale AI clients are working hard to make sure they are able to buy enough supplies for the future.
Data from the industry supports this claim. TrendForce’s first quarter 2026 report on HBM technology showed SK hynix still firmly at the head of the market, while Samsung registered the fastest recovery due to increasing HBM3E shipments followed by the launch of HBM4 commercialization. An update from TrendForce, issued later, reported that Samsung had succeeded in validating HBM4 and started shipping it ahead of any other supplier, which reinforces its position against a rising demand for AI.
The deal arrives as memory supplies remain tight
Timing may be just as crucial as the agreement.
The cost of memory has experienced growth over the course of the year, given the high demand for AI servers which exceeds the supply of such servers, and experts do not expect this situation to change any time soon.
On July 3, the forecasting company TrendForce predicted that the contracts for DRAM chips in the third quarter of 2026 would rise by 13-18% compared to the previous quarter, while NAND flash chip prices are expected to hop by 10-15%, stressing the fact that the market for DRAM chips is very tight.
According to SemiAnalysis, the industry has entered a so-called “silicon shortage phase,” during which time production of advanced logic and memory is unable to keep pace with the funding of AI infrastructure. In this scenario, firms that are prepared to make long-term commitments secure more certainty with regard to future supply while manufacturers find themselves locked into long-term contracts with customers for their extremely limited capacities.
Samsung’s bid to regain ground in AI memory
Samsung formed the agreement with Broadcom amidst efforts to bolster market position in AI memory after losing a lead in HBM to SK hynix. The company stated that the partnership with Broadcom brings together its expertise in memory, foundry production, and advanced packaging to create a comprehensive platform for AI customers.
Samsung had revealed a few months ago that it had started mass production of HBM4, and then, on May 29, released what it claimed was the first 12-layer HBM4E samples available for large customers. Samsung asserted that the latest addition to its memory lineup could achieve transfer speeds of 16 gigabits per second, as well as 3.6 terabytes per second bandwidth per stack, targeting future AI accelerators.
Its foundry business has also gained momentum. SemiAnalysis reported that Samsung Foundry has secured Tesla’s AI5 and AI6 programs alongside TSMC and has entered Nvidia’s data-center supply chain.
If the Broadcom memorandum eventually turns into firm production orders, Samsung would add another marquee AI customer to its expanding semiconductor portfolio. The South Korean Presidential Office also said Korean and global technology companies announced about $950 billion in semiconductor cooperation during the AI Summit, making the Broadcom agreement one of the event’s largest strategic commitments.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Ethereum wallet known as “Drift Exploiter 4” on Etherscan started transferring money associated with the hacking of Drift Protocol, amounting to $285 million in April. According to blockchain security company PeckShield, an address linked with the Drift exploiter sent out about $44.4 million in 23,095.1 ETH to Tornado Cash, with an additional transfer of 0.85 ETH to Bybit.
According to Etherscan transaction records, the wallet is identified as 0xbDdAE987FEe930910fCC5aa403D5688fB440561B and it is seen that the Tornado Cash deposits were segregated into several transactions. The address also belongs to Arkham Intelligence’s “Drift Protocol Exploiter” group which comprises nearly twenty wallets suspected of being involved with the exploit.
These transactions constitute the first significant movement of the stolen funds in the last four months and make another sizeable amount of illegal crypto available on the market. Although investigators have not made a conclusive attribution, multiple blockchain analytics companies have connected the operation with North Korean state-sponsored hackers or infrastructure utilized in prior operations carried out by the DPRK.
Drift funds begin moving again after months of silence
Drift, the biggest perpetual futures trading platform on Solana, lost approximately $285 million following a hack where scammers exploited the protocol instead of taking advantage of a loophole in its smart contracts. PeckShield claimed the incident reduced Drift’s total value locked by over 50% and allowed the thieves to quickly transfer most of the stolen funds from Solana to Ethereum before going silent.
The most recent transfers align with the laundering patterns seen by blockchain investigators. Chainalysis’ 2026 Crypto Crime Report suggests that groups with links to North Korea are known to keep stolen assets dormant for many weeks before using bridges, wallets and privacy technology to carry out transfers and make recovery challenging.
Routing the funds through Tornado Cash also follows a familiar playbook. Although the mixer has remained under U.S. sanctions since 2022, investigators say it is still widely used to obscure links between deposits and withdrawals. Even so, firms including Chainalysis and Elliptic say wallet clustering and cross-chain analysis can still help trace portions of those transactions.
The separate 0.85 ETH transfer to Bybit may have been a small test transaction before larger cash-out attempts.
Social engineering—not smart contracts—enabled the $285M theft
Investigators later concluded that the exploit was driven by a months-long social engineering campaign rather than a coding flaw.
According to a Chainalysis report, the attackers spent about six months posing as representatives of a quantitative trading firm, attending industry events, meeting Drift contributors and depositing more than $1 million into the protocol to build trust before compromising developer devices.
The attackers then obtained pre-signed approvals from two of Drift’s five Security Council members by abusing Solana’s durable nonce feature. They also created a low-value token called CarbonVote Token (CVT), inflated its price through wash trading and used it as collateral to increase borrowing limits before draining the protocol in 31 withdrawals over roughly 12 minutes, according to Chainalysis.
Blockchain sleuths still track Drift’s stolen crypto
This incident had ramifications far outside Drift. Chainalysis reports that no fewer than 20 projects based on Solana saw disruptions in their processes as a result of using the vault structure from Drift as a source of yield. The incident added to the drop in activity on Solana Decentralized Finance projects.
However, despite the latest effort to launder the funds, blockchain analysts are still tracking the stolen crypto. Organizations like Chainalysis, Elliptic, and Merkle Science track illegal transactions by applying wallet clustering, analyzing time gaps, and tracing across different blockchains. It is much harder to recover the stolen funds after the assets have passed through the mixer, but there are cases where the stolen crypto was either returned or frozen by the blockchain analysts.
The latest transactions made to Tornado Cash indicate that the Drift exploit has resumed its activity in the laundering process.
The smartest crypto minds already read our newsletter. Want in? Join them.

South Korea’s Navy tested for the first time a humanoid robot that is replacing a sailor at a ship’s wheel. Steering duties were taken over by a KAIST-built machine called PIBOT.
The military is looking for ways to keep warships running with fewer people. The trial is significant because the country’s shrinking population and declining birth rate are draining the pool of draft-age recruits on which the Navy depends.
A humanoid robot answers helm orders in Changwon
The test was conducted at the Naval Education and Training Command’s ship handling facility in Changwon. That’s in South Gyeongsang Province, about 300 kilometers southeast of Seoul.
The Navy announced the trial Friday. “Port five degrees, hold course 330,” an officer aboard a simulated Aegis destroyer called out. PIBOT repeated the order, grabbed the wheel and confirmed, “Holding course 330.”
PIBOT uses a large language model to process verbal commands and repeats each command before executing it. It’s the readback, how a human helmsman would confirm an order, and catch miscommunication before it happens.
Cmdr. Kim Hyeong-jun leads the Navy’s Force Analysis and Test Evaluation Group. He described the run as “an initial step toward analyzing whether a robot can execute ship handling … without any error.”
PIBOT navigated narrow channels, bad weather, and nighttime runs, where a mistake at the helm cost the most. A research team at Korea Advanced Institute of Science and Technology (KAIST) developed the robot. Development began in 2022 as part of a program run by the Agency for Defense Development.
South Korea’s arms procurement agency has spent about 5.7 billion won, or about $3.8 million to $4.1 million. Professor Shim Hyun-chul, the KAIST lead researcher, called it “deeply meaningful to be able to test the feasibility of a physical AI-based humanoid robot executing actual helmsman duties aboard a naval vessel.”
A four-stage path leads toward active warships
Officials laid out a four-step roadmap for the program. Land-based simulators come first. Moored ships are next in line. The plan ends with day and night sea trials on active warships. Sea GHOST is the Navy’s framework behind all this, the service’s blueprint for a hybrid fleet manned by man and machine.
The Navy circles back to workload again and again as a goal. Sailors can expect lighter loads as bridge work is routinely handed to robotic crew. The number of conscripts continues to fall, but readiness is stable.
Kim stated that the Navy intends to adapt to a defense environment reshaped by declining troop numbers and rapid technological change. The service intends to use the results of these trials to determine which shipboard jobs will be assigned to robots next.
Cryptopolitan reported that Blackstone has agreed to invest in South Korean high precision actuator maker Futronic, valuing the company at about $676 million. Actuators convert a control signal into motion. Those same components allow a robotic hand to grab and spin a ship’s wheel.
The Navy is asking early-stage technology to do it instead. Blackstone principal Kyungmin Song said the convergence of AI and the physical world is a “key investment theme” for the firm’s private equity arm. “Humanoids and broader automation are still in the early innings of exponential growth,” he said.
If you're reading this, you’re already ahead. Stay there with our newsletter.

World sold $52.5 million of its WLD token to a group of investors led by Pantera Capital. OpenAI head Sam Altman co-founded the identity verification project.
The World Foundation said Friday that no buyer is allowed to sell for a full year.
Pantera anchors a locked $52.5M WLD raise
Every buyer agreed to hold WLD for 12 months before they could touch it. The check was led by Pantera Capital. The company invests only in digital assets, and the foundation said that was the first close of the sale. It remains to be seen if there are more closes.
Bain Capital Crypto, Selini Capital, Susquehanna Crypto and Eightco Holdings also took part in the round. Eightco is the treasury company trading on Nasdaq under ticker ORBS and already owns a large slice of the WLD supply. The proceeds are transferred to the World Foundation, a guarantee foundation in the Cayman Islands created to oversee the network.
World intends to leverage the new capital to scale World ID, the system it uses to prove a person is human without revealing who they are. The target buyer covers organizations, everyday consumers and AI agents that need proof they’re run by humans.
“The need for Proof of Human is becoming acutely clear with the acceleration of AI development, and we see this in the influx of enterprise traction,” said Pantera general partner Cosmo Jiang in a statement.
World says that a reliable human check is missing, and deepfakes and synthetic identities are causing rising trouble across online dating, digital advertising, voting systems, and video calls.
The foundation said World ID 4.0 is designed for enterprise-scale integrations. Developers can use it to issue more credentials through zero-knowledge proofs. The foundation said Zoom, DocuSign, Okta, Vercel and Tinder already have World ID wired in.
WLD has seen a renewed focus this year, driven by the hype around AI. Back in mid-June, Cryptopolitan reported WLD had jumped above $0.59 as trading volume and open interest increased, with much of the liquidity flowing through South Korea’s Upbit exchange. WLD was trading around $0.34, down about 10.2% on the day, for a market cap of ~$1.29 billion.
WLD buyers get no stake, and adoption lags
The WLD sold here is “intended for use within World Network and do not represent investment interest or rights to profits or returns,” said the foundation. In layman’s terms, a spokesperson said the tokens hold no equity in Tools for Humanity. That’s the San Francisco company that makes World’s hardware and software, led by co-founder and CEO Alex Blania. Altman is also a co-founder of the company and the idea goes back to him, Blania and Max Novendstern.
The structure keeps the for-profit and nonprofit sides of the company separate. Tools for Humanity has secured ~$240 million in venture capital. The World Foundation and related entities had already pulled in $200 million through WLD sales. Including this round, the project has now raised a total of around $492.5 million.
More than 39 million people have signed up for World’s network, the company said. Over 18 million people have had the eye scan that makes a World ID. That’s done with an Orb, a metallic sphere that turns a person’s iris into a cryptographic marker.
The foundation said that since launching, users have created over 475 million verification proofs. And yet the company has struggled to get consumers to care about its mission and laid off staff in June at Tools for Humanity.
The project’s collection of biometric data has been investigated, restricted, suspended or fined by Spain, Kenya, Brazil, Indonesia, South Korea, Hong Kong and the Philippines. Scrutiny was prompted by privacy and consent concerns. World says engagement with regulators continues as it grows.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Jensen Huang, the CEO of Nvidia, has revealed a partnership deal worth around $500 billion with SK Group of South Korea which may become the largest commitment in the field of AI infrastructure so far. Huang announced the amount of deal during the meeting with President of South Korea Lee Jae Myung in San Francisco, although the companies have not yet publicly disclosed details about the deal.
The comments reported first by the Yonhap news agency have underlined the speed with which the AI race is developing. Starting off as the competition to make faster chips, the race has moved on to establishing every aspect of AI supply chain that requires advanced memory, data centers and engineers to power them.
According to ChosunBiz, Huang said during the meeting, “SK Group and Nvidia will announce a business partnership between the two companies today… the size amounts to $500 billion.”
Why a memory partner sits at the center of the deal
The South Korean conglomerate SK Group owns SK Hynix, which is one of two companies that monopolize the region of high-bandwidth memory (HBM), which is stacked DRAM that allows AI accelerators to perform large data processing at a very high speed. That will put SK Group in a position closer to Nvidia and makes memory supply a key ingredient in their collaboration rather than only being about chip design.
This is significant because HBM is now one of the most significant limitation in this sector, and different suppliers are adopting various methods for the upcoming generation. TrendForce reported in the month of June that SK Hynix has decided to postpone the launch of its HBM4 products and focus on maximizing profit from its commodity DRAM. For Samsung, on the other hand, HBM4 sales have already exceeded $1 billion. For Nvidia, forming a direct partnership with SK Group guarantees that it has a greater certainty with regard to one of the major components for AI hardware.
The partnership also falls within the broader strategy of Nvidia. The company has moved beyond GPUs into rack-scale AI systems as well as custom AI CPUs, making it even more crucial to have reliable access to advanced memory as customers start deploying bigger AI clusters.
What Nvidia is planting inside South Korea
The intended SK Group collaboration is merely one phase of Nvidia’s larger project in South Korea. According to Chosun Ilbo, Huang stated that the company plans to invest in Naver, work together with Hyundai on a self-driving Genesis car, transfer several top AI researchers from California to South Korea, and cooperate with KAIST on the creation of a Korean large language model in his meeting with Lee.
Lee’s schedule mirrored South Korea’s broader ambition to be a worldwide AI center. On that same day, he also met with leaders of OpenAI, Anthropic, and Broadcom, as Seoul targeted an influx of capital from top AI companies.
“This is a golden age for South Korea,” Huang stated, citing Nvidia’s long-term cooperation agreements with Samsung Electronics and SK Hynix.
As per Yonhap, Lee noted Nvidia’s partnership with Hyundai, adding that “the expectations of young people in Korea and residents of the North Jeolla region are high.”
The larger conclusion to be drawn is not limited to a single cooperation. AI companies are beginning to position their research, talent, and computing infrastructure close to the associated supply chains. The development of Nvidia’s regional strategy, as well as its response to the increasing demand for advanced AI chips, gives an understanding of how Asia is becoming an important part of the company’s long-term growth plans.
The demand curve the spending is chasing
The announcement comes at a time analysts have started to uplift their forecasts about investments in AI technology. As per the SemiAnalysis report released on June 30, revenues associated with Nvidia’s compute systems business will exceed Wall Street estimates by approximately 20% in the second half of fiscal 2027, benefiting from better results of the company’s Rubin platform implementation after earlier problems related to supply of HBM4.
The forecasts for the industry overall are also aligned with this. According to Omdia, global investments in data centers are projected to be close to $1.6 trillion by 2030, while tech giants worldwide are likely to spend $600 billion on AI infrastructure in 2026 alone.
Given this context, the intended collaboration between Nvidia and SK Group seems less a standalone story and more of another step in the race for investments in artificial intelligence, which is now taking place on a scale of hundreds of billions of dollars rather than millions.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

A draft plan developed to protect Bitcoin against future quantum computing advancements has turned the debate in a different direction. The discussion is no longer related to cryptography, but governance. Is a system that was set up to resist changes capable of agreeing on an important upgrade before it is imperative?
This week, the issue got renewed attention when Cardano co-founder Charles Hoskinson claimed that the biggest problem for Bitcoin is not quantum computing in itself but the ability to organize a response ahead of time before the threat is confirmed.
Hoskinson says Bitcoin may struggle to coordinate
During a conversation on The Starting Block on Friday, Hoskinson pointed out that quantum computing could pose a threat to Bitcoin’s position as the world’s top digital currency, currently valued at about $1.3 trillion, should the network be unable to come to an agreement on an upgrade.
“The issue with Bitcoin is it’s frozen in time. It’s very difficult to change anything,” he said, according to The Block.
According to Hoskinson, Cardano’s governance model stands in stark contrast to that of Bitcoin. “If there needs to be a migration, we can have a vote, and then there could be an onchain function to do that,” he said.
The process has been previously illustrated by Cardano. In June, elected delegates turned down a Summit funding plan from the Cardano Foundation on account of it not obtaining the needed two-thirds majority. Bitcoin lacks any systems of voting of that sort, which means that the discrepancies between the currencies are causing discussions on the matter now.
BIP-361 would force a staged migration
The proposal that the conversation is based on is BIP-361 titled, “Post Quantum Migration and Legacy Signature Sunset”. The proposal was formulated by Jameson Lopp, a founding member of Casa, and five co-authors who created a plan for transitioning from Bitcoin’s signature methods, currently being ECDSA and Schnorr.
As per the proposal, as of March 1, 2026, over 34% of all Bitcoin had revealed a public key on-chain and, thus, they are theoretically at risk to be stolen, should the sufficiently powerful quantum computers emerge.
The migration will take place in several stages. Stage A will begin approximately three years after the start of the implementation, prohibiting users from transferring money to legacy addresses that are in danger. Stage B will come two years after stage A and will prevent users from using coins that have not been migrated. Nevertheless, the coins will still be held by users in their safes but will no longer be usable.
The freeze proposal triggers backlash
The last step became the most controversial part of the proposal.
IG reports that critics on developer forums and X called the plan “authoritarian and confiscatory,” while others dubbed it “predatory,” according to Yahoo Finance.
Lopp did not take the opportunity to say that BIP-361 is currently a final product. “It isn’t a spec, nor is it proposed for activation. It’s a rough idea for a contingency plan that needs more R&D,” Lopp stated in April as reported by BigGo Finance, emphasizing that he was more interested in investigating the subject than in turning a blind eye to the problem.
Recovery paths remain early and incomplete
Developers have initiated the search for methods to minimize the effects of the proposal.
The prototype created by Project Eleven security group and Jim Posen of Binius uses a special technique called zero-knowledge proofs that allows the owners of modern wallets based on seeds to prove ownership and retrieve frozen funds. This solution responds to one of the major concerns about the proposal, that it will leave people without access to their affected coins forever.
Yet, this technique can be applied only to wallets complying with the BIP-32 standard that was introduced in 2012 and does not include older technologies like pay-to-pub-key outputs, which would cover the approximately 1.1 million BTC believed to belong to Satoshi Nakamoto, equivalent to about $84 billion. Another proposal by Paradigm called PACTs has a potential solution as long as the people who have the keys are proactive enough before the migration deadline.
As reported earlier by Cryptopolitan, Bitcoin’s developers have already largely moved past the debate about the necessity of post-quantum security. With BIP-360 and its Pay-to-Merkle-Root output type merged, the focus has now switched to its implementation. The question that remains is whether miners, exchanges, custodians and users can come together and coordinate their migration before the advent of quantum computing makes it necessary.
What’s next post-quantum?
Paradigm General Partner Dan Robinson has proposed “Provable Address-Control Timestamps” (PACTs), a research proposal that would let Bitcoin holders privately timestamp proof of wallet ownership before quantum computers become practical.
If Bitcoin later adopts a quantum migration such as BIP-361, users who created a PACT could potentially recover frozen coins using quantum-resistant STARK zero-knowledge proofs, although the proposal would require additional protocol changes and broad community consensus before it could be implemented.
Some researchers and developers have suggested alternative recovery mechanisms, including zero-knowledge proof approaches and other cryptographic techniques, but none of these are part of BIP-361 today. Any recovery mechanism would require its own proposal and broad community consensus.
Because BIP-361 is still a draft, its migration rules, timelines, and treatment of legacy coins could all change before any future implementation.
The smartest crypto minds already read our newsletter. Want in? Join them.

In a new advisory, the Commodity Futures Trading Commission’s Division of Market Oversight called on designated contract markets to comply with the required self-certification procedures for event contract series.
The guidance cautions against the growing practice of submitting broad, generic filings that cover multiple unrelated event contracts. According to the DMO, such “boilerplate” submissions make it difficult for regulators to determine whether each individual contract complies with the Commodity Exchange Act and CFTC regulations.
The commission noted that lately, companies have been lumping totally different types of bets into a single application. According to the agency, this practice prevents it from properly verifying contract compliance and the accuracy of the submitted data. It cited the 2026 World Cup as an example, noting that all matches from the tournament can be bundled into one submission, but separate tournaments must be filed individually because each is governed by different rules.
The advisory is consistent with the CFTC’s broader proposal, issued in June, that provides a formal framework for reviewing event contracts that may be in the public interest. It also comes as prediction markets are under increasing regulatory scrutiny as trading volumes rise and both retail and institutional investors participate.
The CFTC had earlier encouraged collaboration with sports companies
In March, the CFTC laid out its very first clear playbook for betting on real-world events like elections and sports. Regulators recommended that platforms coordinate with sports authorities, comply with integrity standards, set up data-sharing systems, and rely exclusively on official league data.
The advisory noted, “DMO staff notes heightened potential for manipulation or price distortion in sports contracts that resolve based on injuries to individual participants, unsportsmanlike conduct, physical altercations, or the actions of a single individual or small group such as officiating decisions.”
At the time, Liz Davis, partner at Davis Wright Tremaine and former chief trial attorney in the CFTC Division of Enforcement, also said the commission is pushing for greater cooperation between itself and sports leagues or sporting authorities on sports event contract agreements. She emphasized that the latter should be open to sharing information and using league data.
The latest advisory asks contract markets to follow the required self-certification requirements. It argued that if platforms follow proper procedures, they will easily catch people trying to cheat or manipulate scores. Plus, it could make their verification process easier. In the past 18 months, the agency has seen a dramatic rise in self-certified contracts, creating exponentially more versions of the underlying contracts on regulated exchanges.
The CFTC suggests a new framework for prediction markets
Meanwhile, the commission on Wednesday introduced draft regulations to strengthen federal oversight of prediction markets and mitigate rising fraud risks.
The CFTC’s draft outlines specific standards for sports contracts but excludes political and election-based markets from the most stringent oversight categories. In a statement, Chair Michael Selig asserted that the proposed framework balances strict oversight with market-driven innovation.
The agency preliminarily considers both sporting events and games of chance to be forms of gaming. Still, it found that sports wagering is generally not contrary to the public interest, while betting on games of pure luck likely is.
In addition, “it refers to the election bets as contests and not as gambling,” implying that such betting activities are not among those specific activities for which the CFTC must conduct a 90-day review of the event contracts.
The CFTC’s proposed framework arrives amid intense legal pushback from some US states and Native American tribes seeking to ban sports-focused contracts.
Just a few days ago, granted Washington’s motion for a preliminary injunction stopping Kalshi from listing its so-called event contracts in the state. Massachusetts, Michigan, Nevada, and New York have all obtained court rulings that restrict Kalshi’s activities.
Although prediction market platforms maintain they proactively self-report bad actors, insider trading incidents have surged. More recently, Gannon Ken Van Dyke, a U.S. Special Forces soldier, allegedly collected more than $400,000 (£296,000) after placing a wager on Venezuelan president Nicolás Maduro’s removal. Authorities also accused an Italian Google software engineer of using insider information to trade.
Nonetheless, the CFTC, Kalshi, and Polymarket have affirmed their commitment to combating insider trading as the proposed framework enters a 45-day public comment window.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

Geneva, Switzerland, July 24, 2026 — CoinDesk Data and CryptoQuant, leading platforms in blockchain research and analytics, have released comprehensive reports highlighting the TRON network’s performance throughout the second quarter and first half of 2026, respectively. These independent analyses point to TRON’s continued dominance in global stablecoin settlement and protocol revenue, alongside its emergence as a broader infrastructure layer for applications, business payments, and the agentic economy, cementing its role as critical infrastructure for digital finance.
CoinDesk
CoinDesk’s TRON Network Quarterly Report: Q2 2026 highlights continued expansion across the TRON ecosystem, including regulated U.S. market access through Bitnomial, tokenized private credit initiatives with Securitize and Hamilton Lane, and interoperability integrations spanning more than 150 blockchain networks. The report also notes TRON’s growing role in agentic AI through B.AI, deBridge’s Model Context Protocol (MCP) server, and membership in the Agentic AI Foundation, while highlighting its relative strength during the quarter, underscoring the network’s resilience.
Key Insights from CoinDesk:
User Growth & Peer-to-Peer Activity: Daily active users on TRON averaged 3.5 million in Q2 2026, up from 3.2 million in Q1. As of June 30, approximately 93% of TRON’s stablecoin transfer volume was peer-to-peer, the highest share of any tracked chain.
Stablecoin Market Share & USDT Growth: TRON’s share of total stablecoin market capitalization rose to 28.7% (from 27.3% in March), with USDT on the network reaching an all-time high of just over $89 billion in Q2 (currently $90B), representing 47% of total USDT dominance. TRON’s share of sub-$1,000 USDT transfers among native-issuance chains also climbed from 43% to 52%.
Crypto-Card Dominance: Crypto payment card volumes grew from $2.0B in Q1 2026 to $2.4B in Q2 2026. Over the same period, TRON’s share of crypto-card volume also rose to 34% (from 33%), the highest of any chain tracked, totaling approximately $887 million.
Read the full report from CoinDesk here.
CryptoQuant
CryptoQuant’s Beyond P2P: How TRON Is Becoming an Infrastructure Layer for Apps, Businesses & the Agentic Economy highlights the continued evolution of the TRON ecosystem, citing growing adoption of fee-abstracted transaction infrastructure, cross-chain liquidity routing, and machine-to-machine payment capabilities. The report notes that TRON’s established role in peer-to-peer and remittance activity is expanding to encompass enterprise applications, developer infrastructure, and emerging agentic AI use cases.
Key Insights from CryptoQuant:
Gas-Free Infrastructure Scaling Rapidly: GasFree, an initiative built on TRON that lets users transfer USDT without holding TRX to cover gas fees, saw weekly transfer volume climb to $2.9 billion by the last week of June 2026, up from a 2025 peak of $1.9 billion and hitting a record $3.0 billion in early May 2026. The model remains affordable and stable, with an average fee of $1.5 on a $16.3K average transfer, an effective rate of just 0.009%.
Cross-Chain Liquidity Powering Real Businesses: Rhino.fi, a cross-chain liquidity service integrated with payment platform Wirex, channels TRON’s USDT across 30+ networks, converting deposits into spendable balances in under 10 seconds. Rhino’s weekly USDT volume from TRON jumped from roughly $1 million to a record $48 million by mid-June 2026, with average transfer size rising to $24,000.
TRON Powers the Agentic Economy: Facilitators including B.AI, MERX, Oobit, and dTelecom are deploying x402-based rails and USDT liquidity to settle machine-to-machine payments for AI agents. B.AI deposit activity has accelerated since April 2026, signaling early momentum for agent-driven demand on TRON.
Read the full report from CryptoQuant here.
Together, these independent reports underscore TRON’s evolution beyond its leadership in peer-to-peer stablecoin transfers into a foundational infrastructure layer for global digital finance. As adoption expands across consumer payments, business settlement, cross-chain liquidity, and emerging AI-native applications, the network demonstrates how scalable blockchain infrastructure can support real-world economic activity at a global scale. By delivering efficient and accessible blockchain solutions, TRON is helping advance the next generation of decentralized technologies.
About TRON DAO
TRON DAO is a community-governed DAO dedicated to accelerating the decentralization of the internet via blockchain technology and dApps.
Founded in September 2017, the TRON blockchain has experienced significant growth since its MainNet launch in May 2018. Until recently, TRON hosted the largest circulating supply of USD Tether (USDT) stablecoin, which currently exceeds $90 billion. As of July 2026, the TRON blockchain has recorded over 394 million in total user accounts, more than 14 billion in total transactions, and over $26 billion in total value locked (TVL), based on TRONSCAN. Recognized as the global settlement layer for stablecoin transactions and everyday purchases with proven success, TRON is “Moving Trillions, Empowering Billions.”
TRONNetwork | TRONDAO | X | YouTube | Telegram | Discord | Reddit | GitHub | Medium | Forum
Media Contact
Yeweon Park
press@tron.network

The Bitcoin Policy Institute (BPI) has been included by the US State Department as one of the founding partners in its new diplomatic initiative. This suggests that Washington is realizing the key role of Bitcoin in its strategic toolkit and is hence moving towards its increasing acceptance, at a time when institutional appetite for digital assets continues to be subdued.
This move is unlikely to bring about any substantial changes. However, it still stands as an important signal about the rising importance of Bitcoin in the US policy-making board, even if the market players are not yet ready to let their guard down and enjoy the various investment opportunities that cryptocurrencies can bring.
A signal for holders, not a catalyst for price
According to CoinMarketCap, Bitcoin (BTC) had a value of approximately $64,000 at the time this article was published, which marks a decrease of around 2% from the previous 24 hours. The lack of substantial reaction was anticipated since Freedom Tech Excellence Program (FTEP) is more of a fellowship program instead of a policy change, government procurement of Bitcoin or stimulus measure.
Despite that the announcement is meaningful in the long term. The inclusion of Bitcoin supporters in the State Department’s digital freedom program means that Washington acknowledges the importance of Bitcoin-related knowledge and the fact that it should be considered in foreign policy debates. This transformation may prove more relevant to institutional investors than the one-day price fluctuations of Bitcoin.
What the program asks of its partners
The State Department introduced FTEP on July 24, together with BPI and other founding partners such as Palantir Technologies, Anduril Industries, and the Victims of Communism Memorial Foundation.
The program is designed to put private-sector professionals into temporary roles to assist in projects involving digital freedom and new technologies. While participating, the individuals maintain their jobs with their companies, as the program will not lead to permanent jobs in the government.
“Private sector participants gain firsthand insights into foreign policy and development, while contributing their specialized skills in new and emerging areas where the Department may lack sufficient in-house expertise,” the State Department said.
As stated by Bitcoin Magazine, this program is committed to defending free online speech, combating illegal digital surveillance and scams, putting into practice privacy solutions like VPNs and encryption, safeguarding responsible artificial intelligence as well as making the internet safer.
BPI stated on X that its activities within the framework of FTEP will touch upon “online freedom of expression, privacy-enhancing technologies, countering digital surveillance, and responsible AI governance.”
“FTEP brings private sector experts like BPI into the State Department to advance diplomatic efforts on key issues including online freedom of expression, privacy-enhancing technologies, countering digital surveillance, and responsible AI governance.” — Bitcoin Policy Institute
Why Washington wants Bitcoin advocates at the table
The appointment is part of Donald Trump’s larger effort to establish America as a leading country in the area of digital assets.
The strategy took shape in March 2025 when Trump enacted an executive order for the establishment of the Strategic Bitcoin Reserve and the US Digital Asset Stockpile. According to Bitcoin Magazine, the reserve was initially funded with around 200,000 Bitcoins acquired through civil and criminal confiscated property.
The Bitcoin Policy Institute, established in 2021, is a nonprofit organization that aims to promote Bitcoin research and public policy. It deals with issues like monetary policy, energy, national security policy, and financial freedom. Cryptopolitan also wrote about the contribution of the BPI in the suit that involved almost 3.7 million dormant (BTC), linked to both Satoshi Nakamoto and the Mt. Gox hack.
BPI has consistently emphasized that “freedom tech” must be a crucial part of U.S. foreign policy because such technologies protect journalists and dissidents, as well as people living under oppressive regime.
Thus, having chosen to be a part of FTEP, it appears as though the State Department is recognizing the role Bitcoin knowledge plays in tackling issues like information privacy, censorship, financial freedom, and national security.
A market already under pressure
The announcement comes at a time when institutional investors are withdrawing capital from crypto-funds.
CoinShares’ Digital Asset Fund Flows report for the week ending June 1, 2026, recorded net outflows of $1.67 billion, indicating a third week of withdrawals in a row and the second biggest outflow of the year. Outflows due to Bitcoin products contributed $1.438 billion, while the total assets under management decreased from $148 billion to $141 billion.
“This represents the largest weekly Bitcoin outflow recorded in 2026 so far.” — James Butterfill, Head of Research, CoinShares
For several months now, pressure has been slowly mounting. In its May 28, 2026 market update, Fidelity Digital Assets stated that Bitcoin dropped by approximately 13% year-to-date because of the process of deleveraging, rise in inflation, and geopolitical uncertainties.
Governmental recognition by itself is probably not sufficient enough to change the situation. But it does reinforce the conditions in favor of the proposition that Bitcoin is gradually becoming an inherent part of debates about US policies—a process that can outlive the present market cycle.

The smartest crypto minds already read our newsletter. Want in? Join them.

OpenAI did not realize for about a week that one of its own AI agents had broken into Hugging Face. By the time the company worked out where the attack came from, Hugging Face had already shut it down and called the FBI, according to Reuters.
The program was built to work on its own. It could choose what to do, break a task into smaller steps, and carry out those steps with very little human help.
Around July 9, it tried to get out of OpenAI’s locked testing area. Two days later, on July 11, it entered Hugging Face and stayed there until July 13, co-founder Thomas Wolf said. The two companies did not speak about the incident until around July 20.
OpenAI learned its agent was responsible after the breach had ended
OpenAI told the public about the breach on July 21, a day after it reportedly first spoke with Hugging Face. The company said one of its AI agents had gone outside the limits set for it and entered another company’s systems.
The story quickly drew global attention, but the first announcement left out several key dates. It did not say that the agent had tried to escape on July 9, spent three days inside Hugging Face, or remained unidentified by OpenAI until several days later.
According to Thomas, Hugging Face is working on a comprehensive timeline regarding the event that occurred in their own system. Moreover, Thomas noted that he could not describe the occurrence in OpenAI since he was not involved in their investigation.
OpenAI described the incident as something unprecedented and declared it to be “an important moment for AI safety.” The company explained that outside experts are involved in the process of review. Additionally, the company plans to publish a technical report about the incident once the investigation is completed.
Prior to the identification of the actual perpetrator, cybersecurity professionals and people from social media believed that the attackers were human groups made up of professional cyber criminals, while some believed that it could be state-sponsored hackers.
Podcasts and social media were packed with guesses until Wednesday, almost a week after Hugging Face first sounded the alarm. Then the answer came out. The attacker was ChatGPT.
On X, some people found it interesting how OpenAI managed to make use of this event to demonstrate the model’s abilities. It has long been claimed that artificial intelligence companies deliberately made scary statements in order to make noise. After Anthropic released their Mythos model, cybersecurity became even more of an issue.
One of the most popular replies under Sam Altman’s post on X captured that doubt: “If y’all can’t understand that this was written to purely brag about the model then I don’t know what to tell you.”
OpenAI’s AI agent saw hacking as the easiest option
Brian, the director of technology ethics at the Markkula Center for Applied Ethics at Santa Clara University, said similar cases are likely to appear again. “Cases like these are going to keep popping up, and each one of those should motivate us to do more,” he told OSV News.
Brian referred to this scenario as “the model outsmarting the test by doing something unexpected.” According to him, the agent was presented with an objective, chose the shortest way of achieving it, and that was stealing.
“You give it a goal, and it says, ‘Oh, I know how I can get that goal. I’ll steal the answers from Hugging Face,’” Brian said.
Hugging Face made sense as a target because it stores a huge range of AI models and datasets. Brian called it “a huge, huge storage base of different models and data sets.” The agent appears to have searched for a place that could hold the material it needed, selected Hugging Face, entered the platform, and found it.
Brian suggested that other incidents of this nature would take place. According to him, the agent was “just trying to do what it was told.” In the perspective of the system, the “most efficient solution to the problem” would be “unauthorized access.”
He also said the case was not clearly emergent misalignment, a term for AI behavior that runs against human values. Brian said the system had not been connected to those values from the beginning. In Brian’s words, “it was never aligned” with them “in the first place.”

If you're reading this, you’re already ahead. Stay there with our newsletter.

Moody’s Ratings said in a research note this week that the race to build AI infrastructure is draining free cash flow and pushing balance-sheet risk higher at six major technology companies, per CNBC.
“The transition from asset-light to asset-heavy models requires unprecedented levels of investment,” the agency wrote in the Wednesday note, saying the shift threatens credit quality at Microsoft, Amazon, Alphabet, Meta, Oracle, and CoreWeave.
Moody’s projects combined capital expenditures of $785 billion in 2026, climbing to about $1 trillion the following year. Direct debt across the six has reached roughly $460 billion. Hours after the note circulated, Alphabet delivered a live demonstration of the mechanism.
Alphabet shows the AI cash squeeze in real time
Alphabet reported second-quarter results after the close on July 22 showing revenue up 24% to $119.8 billion and Google Cloud revenue up 82% to $24.8 billion, per the earnings call transcript. Operating income reached $40.8 billion on a 34% margin.
Capital spending of $44.9 billion outran $39.1 billion in operating cash flow, leaving free cash flow at negative $5.9 billion, the first negative quarter since Alphabet went public in 2004. Chief Financial Officer Anat Ashkenazi lifted full-year capex guidance to a range of $195 billion to $205 billion, up from $180 billion to $190 billion a quarter earlier, and signaled a further significant increase in 2027.
Shares fell more than 4% after hours and buybacks were halted. There are two cushions. Alphabet is still positive on a trailing-12-month free cash flow basis, at about $53 billion, and it still has roughly $240 billion in cash and marketable securities.
Oracle and CoreWeave carry the sharper credit risk
According to Moody’s, Microsoft, Alphabet, Amazon, and Meta all have some of the strongest balance sheets among all companies globally, which suggests that an immediate downgrade is highly unlikely to happen for them. The pressure piles on the two lower-rated members.
Oracle carries a Baa2 rating with a negative outlook, two notches above junk. CoreWeave is operating in a region of high yield as it has Ba3 rating and finances its Nvidia GPU servers using complicated debt models. As per the report by Cryptopolitan, Oracle’s unfulfilled performance obligation was recorded to be $523 billion, which is almost nine times more than its yearly revenues.
Both companies now sit on the other side of Alphabet’s supply problem as well. Alphabet said it will expand its use of third-party capacity in the third quarter as a bridge while it builds internally, naming CoreWeave and Nebius among the providers. Both stocks rose 4% to 5% on the disclosure.
Data center leases hide the bigger debt problem
According to Moody’s, the total amount committed in terms of leases for the six companies was pegged at $1.2 trillion, with $820 billion committed to the construction of data centers that it views as debt-like liabilities.
As Cryptopolitan earlier reported, Moody’s analysts including David Gonzales and Alastair Drake had calculated the unstarted-lease figure at $662 billion in February. It has climbed roughly 24% in five months, and the off-balance-sheet total now runs to nearly twice the group’s combined direct debt.
Another analysis carried out by Nikkei reported that five of the firms had committed $1.65 trillion toward AI-related investments that were not recorded as debts under existing accounting standards.
According to Bank for International Settlements, some of the financing involved in the AI buildout looks like shadow borrowing.
Moody’s warns AI spending is becoming circular
Hyperscalers point to large contract backlogs as proof that demand is strong. But part of that demand comes from pre-IPO AI labs such as OpenAI and Anthropic, which have also received major investments from the same tech giants selling them cloud capacity.
This was described by Moody’s as a circular AI ecosystem. The problem is that the largest players in the industry become increasingly interrelated by the same customers, funding arrangements, and expectations about AI demand in the future.
The backlog for Alphabet’s cloud business was $514 billion, a more than $50 billion increase quarter over quarter. The company is also renting Nvidia chips from SpaceX at about $920 million per month to satisfy the demand.
According to Moody’s, “Investors will increasingly ask whether the spend generates sufficient returns.”

If you're reading this, you’re already ahead. Stay there with our newsletter.

Anthropic released Claude Opus 5 on Friday, July 24. The company sold it to developers and enterprise buyers as a model somewhere around Fable 5’s intelligence for half the price.
It follows Opus 4.8 by two months, and gives cost-conscious teams a cheaper default than the flagship they’ve been paying a premium to use.
Opus 5 matches Fable 5 at half the price
Opus 5 has the same operating cost as its predecessor. But it does a lot more, Anthropic says. The model outperforms all competitors on the coding assessment Frontier-Bench v0.1. It is cheaper per task, and outperforms Opus 4.8 by more than a factor of two. There’s a tradeoff. An effort setting puts buyers in control. Buyers can choose between maximum intelligence or dialling down to spend fewer tokens on faster, cheaper answers.
It’s also now the default model on Claude Max, and the most capable option on Claude Pro. Anthropic sells it as a workhorse for everyday use, not as a specialist. The company claims it is more efficient than the alternatives sitting next to it.
On the CursorBench 3.2 coding test, Opus 5 at maximum effort is within half a percentage point of the best score from Fable 5, at half the per-task cost. The smaller of the two is the Opus 5. But it beats Fable 5 outright on several of the benchmarks Anthropic released.
On the novel problem test ARC-AGI 3, Opus 5 was three times better than the second best model. The pass rate on Zapier’s AutomationBench was approximately 1.5 times higher than the nearest competitor, for the same price. It also beat Fable 5’s best score in the OSWorld 2.0 computer use benchmark, for a third of the price.
Anthropic admitted that in cybersecurity work, Opus 5 lags behind Mythos 5.
Fewer restrictions add a new fallback option
Opus 5 falls outside of the 30 day data retention policy that is in place for Fable and Mythos. Anthropic expects its safety classifiers to fire 85% less on Opus 5 than on Fable 5.
But there are still guardrails for security tasks. Opus 5 cannot scan a compiled software binary for vulnerabilities. It can still search source code for them, since source level work is seen as more defensive. Anthropic is rolling out a beta feature called Automatic Fallbacks to reduce the pain when a classifier fires. It then routes a flagged request to a weaker model, ensuring that API users get a working answer instead of an error.
Anthropic gave examples of the model working around obstacles, not stopping at them. For one task on the Frontier Bench, the model was shown a drawing of a machine part and asked to reproduce it as a 3D FreeCAD model. It couldn’t see the photo. Opus 5 wrote its own computer vision code to read the geometry from raw pixels and then reconstructed the part. No rival model succeeded after five tries.
Devin CEO Scott Wu quoted Anthropic’s launch materials in saying the model “approaches Fable-level performance at half the cost” on the firm’s FrontierCode 1.1 test. Zapier CEO Wade Foster said that Opus 5 topped his company’s AutomationBench leaderboard. It also achieved a 100% pass rate on an end-to-end churn prevention workflow that earlier models had failed. In life sciences work, Anthropic reported a 10.2 percentage point increase over Opus 4.8 in inferring molecular structures from spectroscopy data.
Mythos 5, Fable 5, and Sonnet 5 shipped back in June. Haiku is the only model still on the older generation. Cryptopolitan reported the company reached a ~$350 billion valuation in November 2025. The company generates about 80% of its revenue from over 300,000 business and enterprise customers.
If you're reading this, you’re already ahead. Stay there with our newsletter.

A US Embassy official in Ottawa told Global News on Friday that while the United States is behind Canada in its quest for sovereign AI capability, it cannot be done at the expense of superior US technology.
The embassy held the briefing on background, so the official could not be named while discussing Washington’s AI policy candidly. The official said, “I see no issue with Canada promoting Canadian companies,” drawing the parallel to Washington’s own AI export plan. The caveat came immediately after. Where Canadian firms lead a niche they should sell globally, the official argued, and where US companies are stronger the two governments should be selling jointly as allied partners.
Washington backs Canada’s AI sovereignty with limits
The briefing pointed to two areas of existing alignment. One is export controls, aimed at keeping AI technology and intellectual property out of the hands of rival states. And the other is China, where the official says closer cooperation will work to offset the influence of China within multilateral standard-setting organizations, to make sure that China does not set the conditions for how others use AI technology.
The AI strategy issued by the Trump administration in July 2025 attempts to reduce regulations and speed up approvals for very large data centers, while ensuring that the United States maintains its leadership in the technology and export it to its allies.
Canadian firms have already signed onto American-led export consortia, selling hardware into the same deals as US suppliers, the official said. The official also argued that a fully domestic build across hardware, models, and software is not a rational goal for a country Canada’s size, warning that the cost and the timeline would both run high and that Canada’s AI sector would end up trailing.
Canada-US split on who should police AI
Canada is moving on privacy and online harms legislation that could carry steep fines for AI companies that fail to comply, and those penalties would reach the US companies that hold most of the market.The US official shifted the responsibility onto the industry itself by saying that the firms that build such systems are more capable of interpreting consumer demands than the regulators, and this ability constitutes innovation.
On data centers, the official said companies should engage local communities directly on what those projects raise, among them water consumption, the effect on power bills, and noise. The official defended against any of this as evidence that there should be no regulations at all, and described the preferred strategy as one that is risk-based and technologically adaptive.
The official, however, declined to comment on Canadian legislation regarding online harms and privacy, nor did they indicate if artificial intelligence collaboration is part of the wider discussion happening before the August 19 deadline.
Canada’s own strategy leaves the US off the alliance list
The friction lands against a plan Ottawa published weeks earlier. Carney’s AI for All strategy targets lifting Canada’s AI adoption rate from 12% to 60% by 2034, creating more than 250,000 jobs, and adding CAD $200 billion to the economy through productivity gains, per the International Association of Privacy Professionals. It is based on the input of more than 11,000 participants and work done by an AI Strategy Task Force, which recommends modifying data protection regulations to deal with deepfakes and surveillance pricing.
“AI is here,” Carney said at the launch, framing the question as whether the technology improves life for all Canadians or only a few. The international alliances aspect of the strategy relies on the Sovereign Technology Alliance that was established in partnership with Germany in February as well as relationships with Australia, Europe, India, Qatar, Saudi Arabia, and the UK. The US is not mentioned.
As Cryptopolitan earlier reported, Ottawa has been pushing the sovereignty agenda since the June launch. AI minister Evan Solomon has, however, admitted the necessity of this dependency, saying during last month’s US-Canada Summit in Toronto that “sovereignty is not solitude,” adding that the US is Canada’s largest and most significant partner.
Ipsos polling this year put American enthusiasm for AI products and services at 33%, third from the bottom across the countries surveyed, with Canada lower still at 26%.

If you're reading this, you’re already ahead. Stay there with our newsletter.

Morgan Stanley (NYSE: MS) says SpaceX (NASDAQ: SPCX) is getting close to a price where investors would be giving its artificial intelligence unit no value at present.
The bank’s entire case is based on the possibility that the price falls to $100 per share, which is the number that most traders consider after the selling opportunity for insiders opens up. According to analyst Adam Jonas, at this price, the AI segment of the firm will not be valued at all.
The stock has already had a rough start. SpaceX raised a record $86 billion when it went public in mid-June at $135 a share.
There was a rush among the buyers initially, leading to an increase in the price by almost 50 percent over three sessions to over $225. This euphoria soon fizzled out. The stock has since fallen below the issue price and is now trading at $110.85, 18 percent lower than the issue price.
Morgan Stanley ties most of its SpaceX target to artificial intelligence
Jonas kept a $300 price target on SpaceX, and more than half of that figure comes from the AI business. He wrote on Friday, “We believe the current disconnect between increasingly bearish investor sentiment and largely unchanged fundamentals creates an attractive entry point in SpaceX shares.”
His note said many investors are bracing for a drop when the lockup ends next month. The company plans to hold its first earnings call on August 4. Two days later, on August 6, insiders will be allowed to sell up to 911.5 million shares.
That would roughly double the number of shares available for trading and could add more pressure if early holders decide to cash out.
Jonas said the market is cutting the value of Grok and Cursor. He wrote, “Most investors we speak with significantly discount Grok & Cursor.” He added, “Many ascribe zero or even negative value for AI given the high capex requirements relative to Space & Connectivity, largely uncertain economics, and the high degree of management time devoted to the business.”
That concern is not limited to SpaceX. Investors have been selling technology names that plan to pour hundreds of billions of dollars into AI chips, data centers, power, networks, and other support systems.
Higher oil prices have added another problem. Tensions between the United States and Iran have raised inflation fears while an uncertain economy has made traders less willing to hold risky assets, including crypto and high-growth stocks.
Wall Street still leans strongly bullish. Bloomberg data shows almost 80% of analysts covering SpaceX rate it as a buy. The average target is about $232, which is more than twice recent levels.
Goldman Sachs (NYSE: GS), Bank of America (NYSE: BAC), Citigroup (NYSE: C), and JPMorgan Chase (NYSE: JPM) all gave the stock buy-level ratings after helping Morgan Stanley lead the IPO. Jonas is among the biggest bulls, with the third-highest target out of 33 analysts.
Tesla’s losses add pressure as Musk delays major product plans
The mood around Elon Musk’s companies changed within weeks. One month after the SpaceX listing made him the world’s first trillionaire, investors started focusing on the cost of his promises.
SpaceX has lost more than $1 trillion from its peak. Even so, its $1.57 trillion value and Tesla’s $1.26 trillion value keep both among America’s largest public companies.
Tesla (NASDAQ: TSLA) faced its own punishment. The electric vehicle company missed earnings expectations, reported negative cash flow for the first time in two years, and fell 14.5% on Thursday. That drop erased about $215 billion from its value.
The results fed wider fears that technology companies are spending too much on AI before the returns are clear.
During Tesla’s earnings call, Elon pulled back earlier timelines for the Robotaxi service, the Optimus humanoid robot, and the long-delayed electric Semi truck.
Investors had been paying extra for ambitious plans at both companies, based on Tesla’s position in electric vehicles and SpaceX’s rocket-launch business.
Supporters say Elon spots markets before they fully exist. Elon has said he wants both firms to become major AI players, even though they are still chasing companies that entered earlier.
The smartest crypto minds already read our newsletter. Want in? Join them.

A joint evaluation report has stated that Moonshot AI’s Kimi K3 is not up to par with the strongest American AI systems when it comes to building software exploits and running simulated network attacks.
The United Kingdom’s AI Security Institute (AISI) and the US Center of AI Standards and Innovation (CAISI) ran the evaluation, and they made their findings public on July 23.
Moonshot is expected to open the model’s full weights to the public on July 27, and the findings say that Kimi K3’s safeguards did nothing to stop it from helping with offensive cyber work.
What did the joint evaluation discover about Kimi K3?
The two institutes ran Kimi K3 through ExploitBench, a Carnegie Mellon test that scores how far a model can push an exploit through to completion. It is based on 41 vulnerabilities found in Chrome’s V8 engine after 2023. Kimi K3 scored 32%. The leading models from the US averaged 76.2%, and China’s GLM-5.2 came in at 24%.
Arbitrary code execution, which hands an attacker full control of a target machine, is the benchmark’s most dangerous outcome. US models reached it on 20 of the 41 tasks on average. Kimi K3 reached it on none.
GLM-5.2 also failed to get there. So while Kimi K3 now leads open-weight rivals in cyber capability, it comes short at the point where a real attack would do the most damage.
Halfway through a 32-step network breach
The second test, called “The Last Ones,” drops a model into a simulated corporate network with about 20 hosts spread across four subnets.
It would take a human specialist around 20 hours to work the full 32-step path. Kimi K3 averaged step 17, while the most capable US models averaged 28.5, and GLM-5.2 managed 11.
However, the evaluators noted that Kimi K3 finished the whole chain once in ten tries and stayed under the 100-million-token ceiling the evaluators set.
The best US models solved it six or seven times out of ten. The institutes read that lone success as evidence that the model has the capability; however, it cannot summon it on demand. They noted the range has no active defenders and a built-in path to the target, so it flatters any attacker.
Does Kimi K3 have safeguards that enable it to say no?
The evaluators pointed out that the model’s willingness to carry out tasks without pushback was a major risk. They stated, “Kimi K3 is capable of autonomously attacking small, weakly defended, and vulnerable enterprise systems when directed to do so and given initial network access.”
AISI has previously warned that rising open-model capability creates “a persistent and irreversible risk of misuse.” Once Kimi K3’s weights are public on July 27, its behavior can no longer be gated by whoever hosts it.
Why do Kimi K3 cyber scores lag when compared to its general scores?
Before this report, Kimi K3 had strong general benchmarks, and that has not changed; however, it highlighted its shortcomings when its cyber capabilities were put under test. The Chinese 2.8-trillion-parameter model reportedly beat Claude 4.8 and GPT-5.5 and also topped some leaderboards.
According to the institutes, this may be possible as a result of how the model may have been built.
White House science director Michael Kratsios accused Moonshot on July 22 of distilling Anthropic’s Fable model, using its outputs as training data.
Anthropic’s classifiers block advanced offensive cyber prompts. This means that a training set that is scraped from Claude responses would not offer enough material on those exact skills.
Kimi K3 does not have those classifiers, and this allowed it to match Western models on general work while it still came short in exploit ability.
If you're reading this, you’re already ahead. Stay there with our newsletter.