Core Scientific’s transformation from bankrupt Bitcoin miner to U.S. artificial intelligence infrastructure provider accelerated during the second quarter of 2026. 

The company reported approximately 1.1 gigawatts of leased customer power capacity and more than $24 billion in potential revenue under long-term contracts.

The July 28 results also showed that high-density colocation had become the company’s largest business. Colocation produced $136.7 million of Core Scientific’s $164.2 million in quarterly revenue. Digital asset self-mining contributed $21.5 million.

The turnaround is supported by signed contracts with CoreWeave and companies deploying AMD technology. However, the $24 billion figure is neither cash received nor guaranteed profit. It represents potential revenue spread across contracts lasting as long as 15 years.

Core Scientific must still construct the facilities, deliver capacity on schedule and keep customers operating under those agreements. Its latest filing also shows that the transition requires heavy spending and considerably more debt.

Core Scientific preserved power assets through bankruptcy

Core Scientific filed for Chapter 11 protection in December 2022 after Bitcoin prices fell, electricity costs rose and financing conditions tightened. The company also faced a payment dispute with bankrupt cryptocurrency lender Celsius, previously one of its mining-hosting customers.

The company had about $4 million in cash when it filed. Its problem was not an absence of physical assets. It had already invested heavily in mining facilities, substations, land, equipment and access to utility power. Those assets could not produce enough cash to meet its immediate liabilities during the market downturn.

A Texas bankruptcy court confirmed the company’s reorganization plan in January 2024. Core Scientific emerged later that month and resumed Nasdaq trading under the CORZ ticker. The restructuring reduced its debt by approximately $400 million through conversions of equipment financing and convertible-note claims into equity.

More importantly, the company retained 724 megawatts of operating capacity across five U.S. states. That infrastructure became valuable as AI companies began competing for sites with secured power, fiber connections and space for high-density computing equipment.

Building a new data center can require years of grid studies, permits and construction. Core Scientific already controlled energized sites built for power-intensive Bitcoin mining. Management began marketing that infrastructure to AI customers seeking faster deployment.

CoreWeave established a new use for mining capacity

Core Scientific’s commercial shift began with CoreWeave. In June 2024, the companies signed 12-year agreements covering about 200 megawatts of high-performance computing infrastructure. Core Scientific estimated more than $3.5 billion of cumulative revenue under the initial contracts.

Several expansions increased CoreWeave’s contracted capacity to approximately 590 megawatts. Core Scientific now associates those agreements with approximately $10.2 billion in potential revenue over their terms.

The contracts also reduced some of Core Scientific’s initial financing burden. CoreWeave agreed to fund parts of the construction, with those amounts credited against future hosting payments. During the first six months of 2026, CoreWeave funded $180.9 million of Core Scientific’s capital expenditures.

As crypto.news previously reported, the company sold $208.3 million of Bitcoin to finance its AI transition during the first quarter. The sale represented a clear change from a model centered on accumulating and mining BTC.

CoreWeave attempted to acquire Core Scientific before the infrastructure transition was fully completed. Core Scientific rejected a $5.75-per-share cash offer in 2024, saying it undervalued the business.

The companies agreed to an all-stock transaction valued at approximately $9 billion in July 2025. Shareholders rejected the transaction on Oct. 30, after opposition focused on the fixed exchange ratio, valuation and exposure to CoreWeave’s share price. The companies then terminated the agreement but retained their commercial relationship.

AMD expands the Core Scientific AI pipeline

Core Scientific announced a wider infrastructure partnership with AMD on July 28, 2026. The arrangement starts with approximately 530 megawatts across five U.S. sites under 15-year agreements.

The company associates that initial capacity with more than $14 billion in potential base contract revenue. Deployments are scheduled to begin in 2027 and will support customers using AMD Instinct accelerators, EPYC processors and ROCm software.

AMD also received reservation rights covering another 1.925 gigawatts. If all reserved capacity becomes contracted, the partnership could reach approximately 2.5 gigawatts.

That larger figure remains conditional. Reservation rights are not the same as executed leases. Development will depend on customer demand, available grid capacity, construction progress and additional financing.

The arrangement gave AMD warrants to purchase as many as 30 million Core Scientific shares at $23.47 each. Approximately 6.5 million warrants vested after related leases were executed in July, according to Core Scientific’s quarterly filing.

Some initial capacity was leased to AI infrastructure operator Neocloud. AMD entered a credit-support arrangement connected to equipment installed for Neocloud, but the filed contracts do not describe AMD as an unconditional guarantor of every Neocloud payment.

The expansion reduces Core Scientific’s reliance on CoreWeave at the contracted-capacity level. Customer concentration nevertheless remains material because a limited number of counterparties support most of the company’s colocation revenue.

AI revenue has overtaken Bitcoin mining

Core Scientific’s financial results show that the transition has advanced beyond announced plans. Second-quarter colocation revenue rose to $136.7 million from $10.6 million one year earlier.

The segment produced approximately $80 million in quarterly gross profit and a margin near 59%. Self-mining produced $21.5 million in revenue but recorded a gross loss of approximately $12.2 million.

Core Scientific said 437 megawatts were generating billable revenue by the end of the quarter. That capacity represented about $635 million in annualized GAAP hosting revenue, based on the company’s calculation.

The business was not profitable under GAAP. Core Scientific reported a quarterly net loss of approximately $1.16 billion and an operating loss of $78.5 million.

About $1.05 billion of the net loss came from fair-value changes involving warrants and contingent value rights. Those expenses were non-cash accounting charges linked mainly to movements in Core Scientific’s share price. They did not represent an equivalent cash payment during the quarter.

Adjusted EBITDA reached $41.1 million. This is a non-GAAP measure that excludes several expenses and should not be treated as equivalent to net income or unrestricted cash flow.

CORZ traded near $18.72 on Aug. 20, giving the company a market capitalization of approximately $6.1 billion. The share price remained several times above its $3.44 closing price on the first day after its January 2024 relisting.

The $24 billion pipeline carries a financing test

Core Scientific spent $954.2 million on property and equipment during the first half of 2026. It also completed a roughly $232.5 million acquisition of land and development rights for a proposed 430-megawatt site in Hunt County, Texas.

The company financed part of its expansion by issuing $3.3 billion of senior secured notes in May. The notes carry a 7.75% interest rate and mature in 2031. Long-term debt reached approximately $4.3 billion by June 30, up from about $1.06 billion at the end of 2025.

Core Scientific held approximately $1.8 billion in cash, cash equivalents and digital assets at quarter-end. However, it expects continued construction spending as it converts former mining sites and develops new campuses.

The company’s first-half operating cash flow also requires context. It benefited from Bitcoin sales, customer construction funding and changes in working capital. Those sources do not show that recurring colocation income can already finance the full development program independently.

Similar economics are encouraging other miners to reuse power infrastructure. In related coverage, crypto.news found that Bitcoin miners are increasingly converting energized sites into AI data centers as mining margins weaken.

Core Scientific’s immediate targets include bringing more CoreWeave capacity online and starting the contracted AMD-related deployments in 2027. Delivery schedules, construction spending, customer performance and additional lease conversions will determine how much of the advertised backlog becomes recognized revenue.

The turnaround rests on a genuine shift in revenue, but its completion remains forward-looking. Core Scientific has exchanged direct exposure to Bitcoin prices and mining difficulty for construction, financing and customer-credit risks. The company’s next test is converting contracted megawatts into reliable, billable computing capacity without allowing its debt burden to outpace operating earnings.

FAQs

Senator Ruben Gallego has urged U.S. lawmakers to slow down the push to put the CLARITY Act on the Senate floor, warning that moving too quickly—before remaining disputes over ethics and stablecoin yield are resolved—could derail the bill’s chances of ultimately clearing Congress.

Speaking at the SALT Wyoming Blockchain Symposium on Wednesday, Gallego said the industry should encourage continued negotiations among Senate Democrats and Republicans rather than forcing an immediate vote. He also pointed to unresolved procedural work needed to package the measure—including committee handling and logistics for sending it to the House.

Key takeaways

• Sen. Ruben Gallego warned that a “fast vote” on the CLARITY Act could produce an outcome that lawmakers do not want, potentially setting the legislation back.
• He said Democratic support depends on establishing “sufficiently strong” ethics restrictions, and that the White House has not provided a detailed response to compromise language.
• The concern aligns with broader pressure from the Trump administration to move toward swift passage, including calls to pass a “fair version” of the bill.
• Gallego’s comments also reflect the practical reality that Senate leadership expects to delay action until September, suggesting negotiations may still be unfinished.

Why Gallego says timing matters for the CLARITY Act

Gallego’s intervention frames the CLARITY Act not just as a policy debate, but as a coalition-building challenge. He argued that legislators still must complete multiple steps before a vote can happen in a way that stands a realistic chance of meeting the Senate’s threshold for passage.

While the Senate could, in theory, take up legislation immediately, Gallego emphasized that lawmakers are still working through outstanding components. He said Congress has to address the bill’s Agriculture Committee portion, assemble the broader package, and determine how the legislation would move to the House.

In his remarks, Gallego urged colleagues not to “go for a fast vote,” explaining that rapid action can lead to a quick result without delivering the desired final outcome. He added that premature movement could “set it back further,” underscoring his view that the political process is still unsettled.

Tillis and Gallego say White House hasn’t answered ethics proposal

A central point in Gallego’s warning was ethics. He said he and Republican Senator Thom Tillis submitted compromise ethics language to the White House ahead of the congressional recess, but that the administration has not responded point-by-point.

Gallego described a pattern of outreach that, in his account, has produced no clear engagement: he said the proposals were repeatedly sent and returned “blank,” returned with language that was not as forward-moving, or were met with no response. He argued that if ethics restrictions are not strong enough, it will be difficult to attract Democratic support—support he views as necessary to move the bill forward.

Cointelegraph attempted to obtain comment from the White House but did not receive a response before publication.

Administration pressure contrasts with Senate procedural delays

Gallego’s comments come as the administration continues to press for progress on the CLARITY Act. Earlier this week, Trump urged Congress to pass a “fair version” of the bill during a White House appearance with crypto executives.

At the same time, Senate leaders have indicated that the chamber will not push for immediate action. Reporting from Cointelegraph said that Senate Majority Leader John Thune confirmed on Aug. 7 that the Senate was “punting” the vote and that the bill would be queued up “first thing” after lawmakers returned from recess—language that effectively sets the focus on a September timeframe rather than an immediate floor push.

White House crypto adviser Patrick Witt previously indicated that the administration would continue negotiating with Democrats until the September vote, while also adding that it “can’t afford to wait forever.” This creates a narrow window in which ethics language and other unresolved elements must be settled enough for negotiators to build a coalition capable of reaching the Senate’s 60-vote threshold.

Gallego’s remarks suggest a tension between that timeline pressure and what he believes still needs to be resolved. In his view, if lawmakers prematurely force a vote before disputes are resolved and the coalition is assembled, the bill risks failing—or winning only in a form that satisfies fewer members than required.

What remains unclear, and what to watch next

Gallego’s warning turns attention to the mechanics of getting from draft policy to final legislation that can actually clear the Senate and proceed to the House. His comments highlight that even if the CLARITY Act is broadly backed in principle, the details—especially around ethics and how stablecoin yield issues are treated—may be decisive for Democratic support.

Going forward, observers should watch whether the White House provides the point-by-point engagement Gallego says it has not yet delivered, and whether Senate leadership’s September timeline is matched by measurable progress in assembling the bipartisan coalition needed for passage. If negotiations remain unresolved, Gallego’s core concern is likely to take center stage: that rushing the process could reduce the odds of a durable legislative outcome rather than improving them.

Hyperliquid’s native token surged more than 20% over 24 hours after United States President Donald Trump said regulators were working on a compliant pathway to make the decentralized trading platform available to American users.

HYPE traded around $62 immediately before Trump’s remarks and subsequently jumped as much as 16% to a 24-hour high of $72.28, according to CoinGecko. It later settled to about $70, up approximately 20% in the last day, with 24-hour trading volume reaching $1.4 billion. 

“I understand that Mike is also working to bring Hyperliquid into the United States in a fully compliant and legal fashion,” Trump said during a Wednesday White House event, referring to Commodity Futures Trading Commission (CFTC) Chair Michael Selig. “Working very hard on that.”

The market reaction shows how the prospect of US access could reprice HYPE and publicly traded companies holding the token. However, neither the CFTC nor Hyperliquid has released a formal proposal explaining how US access would work, whether an application has been submitted or when a compliant service could launch.

Hyperliquid’s 24-hour price chart. Source: CoinGecko

$65,000 options bet on HYPE treasury firm raises eyebrows

Meanwhile, shares of Hyperliquid Strategies, a Nasdaq-listed HYPE treasury company trading under the ticker PURR, closed Wednesday at $9.39, up 30.4%, according to Yahoo Finance. Despite sharing the protocol’s name, the company said it is independent and not affiliated with Hyperliquid.

Roughly four hours before Trump spoke, someone reportedly paid about $65,000 for 719 PURR call options with an $8 strike price expiring in mid-October, according to CNBC. The contracts were purchased for approximately $0.90 each and were quoted at $2.45 by the close, valuing the position at roughly $176,000 and producing an unrealized gain of about $111,000.

Related: Rushed CLARITY Act vote could set legislation back, Gallego warns

Delayed market data derived from the Options Price Reporting Authority corroborates the unusually heavy activity in the contract. OptiView data showed 2,575 of the October $8 calls traded during the session, compared with just 67 contracts in open interest beforehand. Volume was more than 140 times the contract’s 30-day average.

The publicly available data confirms elevated trading but does not independently identify the buyer or establish that the reported 719-contract order was based on nonpublic information. There is no clear evidence of insider trading, and the CFTC had previously publicly disclosed a July 15 meeting with Hyperliquid Labs and Hyperliquid Strategies.

Magazine: ‘Fabricated rumors’ about BitMart founder, Binance bStocks dominate: Asia Express

Swift, HSBC and Standard Chartered completed the first live interbank transaction on Swift’s blockchain-based ledger on Aug. 19, 2026.

The transaction connected two independently operated tokenized deposit systems through a shared coordination layer.

HSBC recorded its resulting obligation through the bank’s Tokenised Deposit Service. Standard Chartered used its own tokenized deposit infrastructure, according to the banks’ joint release.

The parties did not disclose the transaction’s value, currency, customers or originating jurisdictions. They also did not announce a commercial launch date.

Swift connected two separate tokenized deposit systems

The banks exchanged payment messages through the Swift blockchain ledger. The system matched and netted their respective obligations before recording the results on each bank’s infrastructure.

The transaction demonstrated that banks do not necessarily need to issue deposits on one shared platform. Swift’s model instead connects separate systems and coordinates the instructions passing between them.

However, the ledger did not complete the final movement of conventional money. Settlement occurred through existing payment systems. That distinction means the transaction tested interoperability and orchestration rather than full onchain settlement.

Swift described the ledger as a layer that can preserve established compliance, credit and operational controls. Its planned availability around the clock could allow banks to process tokenized deposit instructions outside normal payment windows.

The Swift blockchain ledger coordinates bank obligations

Tokenized deposits represent claims against issuing commercial banks. They differ from stablecoins, which are generally issued by specialist companies and backed by separate reserve portfolios.

Under Swift’s design, HSBC and Standard Chartered retain control of their own deposit liabilities. The shared ledger coordinates corresponding obligations without creating a new public payment token.

Lewis Sun, HSBC’s global head of domestic and emerging payments, called the transaction a “landmark moment.” That assessment reflects HSBC’s view and does not establish that the system is ready for wider commercial use.

Standard Chartered Global Head of Virtual Accounts and Clearing Mark Willis said the transaction represented a step toward “more seamless, always-on financial services.” Wider adoption will depend on further testing, sufficient liquidity and support across currencies and jurisdictions.

HSBC brings a six-market network into the pilot

HSBC’s Tokenised Deposit Service is live in Hong Kong, Singapore, Luxembourg, the U.K., the U.S. and the United Arab Emirates. It supports the offshore Chinese yuan, Hong Kong dollar, Singapore dollar, euro, British pound, U.S. dollar and UAE dirham.

The bank introduced the service for eligible U.S. corporate and institutional customers in April 2026. HSBC said clients can use it for domestic and cross-border transfers at any time, subject to availability and regulatory requirements.

Standard Chartered operates across 55 markets. The bank did not specify which parts of its network participated in this transaction or where its tokenized obligations were recorded.

Seventeen banks will test wider interoperability next

Swift declared its blockchain ledger ready for initial use on July 9 after nine months of development. Seventeen banks across six continents joined the rollout, as crypto.news previously reported.

The group includes ANZ, BNP Paribas, BNY, Citi, DBS, HSBC, MUFG, Standard Chartered, UBS and Wells Fargo. Other participants include First Abu Dhabi Bank, FirstRand, Itaú Unibanco, Lloyds, Mashreq, OCBC and UOB.

The first transaction provides a live reference point, but it does not confirm production-scale adoption. Swift has not published transaction-volume targets or a deadline for expanding the ledger beyond its controlled rollout.

The next phase is expected to test additional institutions, currencies and operating conditions. Banks will also need to assess liquidity management, reconciliation, compliance checks and the treatment of transactions initiated outside conventional settlement hours.

Bancel says that Moderna is also exploring using intismeran alone, perhaps in earlier stage cancers. For the current trial, because intismeran is still an experimental therapy under study, and an approved therapy, Keytruda, does exist for melanoma patients, regulators at the FDA were more comfortable with combining the drugs so every patient received at least the standard of care and therefore would not be additionally harmed by joining the study. “It would be unethical for us to tell people to come on a study of something that we don’t know if it works, when people are fighting for their lives,” he says. “That’s why we did the study together with Keytruda, which can improve outcomes, and we tried to show that the combination was better, and we did.”

Bancel’s teams are already studying intismeran in stage 1 lung cancer, to see if using it earlier, and priming the immune system, can lead to equally beneficial outcomes for those people. In lung cancer, standard treatment involves surgery and careful monitoring for any signs of returning cancer, so doctors can ethically study how adding intismeran for some patients affects their cancer outcomes compared to those not receiving it. While that study is still in early stages, Bancel is optimistic since “from a scientific standpoint there is no scientific sense that it would work in melanoma and not work in another tumor type,” he says of the mRNA-based approach. “I think it will be transformational. Think about if you get a screening X-ray and find stage 1 lung cancer. Then you get intismeran, which is like a vaccine so you have no severe toxicity like with other immunotherapy. Think about getting this product on your way to work, or even [potentially] at your local pharmacy—that would be an incredible change in care. And it can help reduce the risk of metastasis, and have a profound impact on patients, as well as prevent people from getting very severe disease.”