Cryptopolitan

Baton Corporation, the UK-registered company behind Solana memecoin launchpad Pump.fun, posted a job listing on June 24 for a Chief Legal Officer (CLO) with a salary range of $1 million to $5 million per year.
Co-founder Alon Cohen promoted the opening, stating that they are searching for someone who can own the company’s regulatory and litigation posture.
Baton is currently fighting a consolidated class action in the Southern District of New York that accuses the platform of selling unregistered securities. The plaintiffs also allege that under a separate RICO theory, Baton is operating a racketeering enterprise.
Is Pump.fun under legal siege?
The job posting itself reveals numbers that Baton has not widely publicized. The listing, hosted on Solana’s job board, states that Pump.fun processes more than $300 million in daily trading volume and generated over $500 million in profit last year with a team of fewer than 100 people.
Pump.fun was geo-blocked from the United Kingdom in 2024 after FCA action, according to a Change.org petition calling for a coordinated European criminal investigation into Baton’s operations.
The same petition cites on-chain analysis from Dune Analytics showing that more than 60% of roughly 4.25 million wallets that traded on Pump.fun lost money, with nearly 1,700 wallets losing over $100,000 each.
What is Pump.fun’s legal strategy to the lawsuits?
The class action was filed by lead plaintiff Diego Aguilar through Wolf Popper LLP, and it alleges that tokens launched through Pump.fun qualify as unregistered securities under Sections 5 and 12(a)(1) of the Securities Act of 1933.
However, it is not the only lawsuit that Pump.fun is facing as a separate complaint was filed on January 16, 2025. This complaint focuses on the PNUT token, a Solana memecoin that plaintiffs say reached a $1 billion market capitalization before falling 89% from its peak.
Brown Rudnick LLP has been the law firm representing Baton on retainer.
The RICO statute allows treble damages, the case is sized at approximately $5.5 billion, and plaintiffs have asked the court to appoint a receiver to take control of Baton’s operations. This means that Baton could suffer immensely should the rulings not go in its favor.
Why pay $5 million for a lawyer
A $5 million base would place the CLO among the highest-paid legal officers in crypto, and it speaks to how seriously Baton is treating its legal risk.
The listing states that the candidate should have at least 10 years of experience, a New York Bar admission, and direct experience managing regulatory investigations and enforcement actions.
The CLO role will be reporting to a general counsel and covers US digital asset regulation (SEC, CFTC, FinCEN, OFAC), UK FCA compliance, EU MiCA obligations, and litigation management across jurisdictions.
The trial date for the Aguilar case has not been set; however, the court docket shows that the most recent filing date is April 14, 2026.
The smartest crypto minds already read our newsletter. Want in? Join them.

360 Security Technology, one of China’s largest cybersecurity firms, has announced two AI-powered security tools with claims that they can rival Anthropic’s Mythos vulnerability-detection system, stating that the American model is a strategic threat Beijing must answer to.
Founder Zhou Hongyi introduced the tools at the ISC.AI 2026 cybersecurity conference in Beijing under the collective name “Yitian Tulong,” a reference to a classic Chinese martial arts novel translated as “Heavenly Sword and Dragon Saber,” according to Reuters.
The firstof the two tools, Tulongfeng, automates software vulnerability discovery. Zhou went ahead to call it “China’s version of Mythos.” The second AI tool, Yitianzhen, handles automated cyber defense and incident response. Together, 360 is positioning them as both shield and sword in an escalating AI-driven security arms race.
Mythos’ foundations
In Anthropic’s April Mythos preview, the AI company stated that it was a system capable of autonomously finding flaws in widely used software. The company also reported at the time that an early version had uncovered “thousands” of serious vulnerabilities across multiple operating systems, web browsers, and other applications.
The system’s dual-use potential triggered an immediate alarm, with cybersecurity experts warning that the same defensive capabilities could significantly arm offensive operations. Washington responded by ordering Anthropic to suspend exports of a less powerful Mythos variant to all foreign destinations and nationals, citing national security.
360 Security pitches agents over compute
Zhou acknowledged that Chinese AI models still trail their American counterparts, saying, “Objectively speaking, domestic models still have a 20%-30% gap in base capability.”
His argument, however, was that China cannot afford to wait for that gap to close. Instead, 360 took an “agent” approach as described by Zhou, with a focus on layering AI models on already existing security expertise, vulnerability databases, and automated tooling. He claimed this combination gives Tulongfeng capabilities equivalent to Mythos despite using less powerful underlying models.
“If Mythos is a top-end chip, what we are building is a complete machine that can run stably, work 24 hours a day, and make fewer mistakes,” Zhou said, according to Reuters. “If the U.S. route is to cultivate a genius hacker, 360’s route is to organise a professional attack-and-defence team.”
360 claimed Tulongfeng had identified 3,432 software vulnerabilities so far, with 105 confirmed by Chinese authorities. Earlier in April, the company reported that its AI-driven methods detected about 1,000 vulnerabilities in systems including Microsoft Office.
China’s potential “one-way transparency” problem
Without domestic equivalents to Mythos, Zhou argued that China faces a “one-way transparency” problem where American entities possess the capacity to probe Chinese software and critical infrastructure, with Chinese firms unable to do the same.
“This kind of powerful weapon that can change the landscape of cyber offence and defence cannot be held only by others,” Zhou said, according to 360’s published transcript.
Zhou sits on China’s top political advisory body, which makes his remarks carry additional political weight. His warnings echo coverage from Chinese state media describing Mythos as demonstrating “unprecedented cyberattack capabilities.”
The announcement arrives amid years of mutual accusations between Washington and Beijing over offensive cyber operations targeting critical infrastructure. The enforcement of tighter U.S. advanced chip exports since 2022 has led to constraints in Chinese AI development; however, this performance gap has recently narrowed.
If you're reading this, you’re already ahead. Stay there with our newsletter.

A humanoid robot was filmed kneeling on a sidewalk, bowing to strangers, and asking them to cover its electricity costs. The $16,000 Unitree G1 humanoid robot was seen in China’s Sichuan province. It was essentially begging for money.
The stunt went viral across Chinese social media last week. It featured a donation plate, a QR code for digital payments, and an LED sign reading “no money to recharge.” The robots accepted digital payments through WeChat Pay and Alipay.
At the time of reporting, no one had claimed responsibility for placing the machine there. But some experts believe the stunt was just a viral tech demo.
Commenters split, China’s robot demos keep going sideways
Social media users turned the beggar robot into a punchline in no time. “Even beggars are being replaced by robots,” one commenter wrote. Others speculated the owner was sitting at home while the machine collected income on their behalf.
Not everyone found it amusing. Some questioned why anyone would donate to a machine when people in genuine need go without help. “If we can’t guarantee basic dignity for our own citizens, why would we treat artificial intelligence any better?” one user asked.
In Xinjiang earlier this year, a G1 robot performing a martial arts routine kicked a child in the stomach during a demo, sending the boy to the ground, as Cryptopolitan previously reported. The robot was remotely controlled at the time, and engineers said it was functioning “as intended.” At the moment, there’s no regulatory framework governing how close bystanders should stand to humanoid robots in China or most other countries.
A separate clip showed one of the machines attempting to dance to Michael Jackson’s Billie Jean before it lost its balance and collapsed on stage.
Wei Zhejia, chairman of Taiwan Semiconductor Manufacturing Company, said during a speech in May that Chinese robots “jump around, bounce about” and are “just for show.” Industry analysts backed that assessment, saying that most Chinese humanoid robots are for entertainment and demonstration, and far from practical commercial use.
Cheap hardware, no safety net
China prioritizes the robotics industry because of an aging population and slowing economic growth, which have pushed policymakers to invest heavily in automation.
Unitree told local media earlier this year that it expects to ship between 10,000 and 20,000 units in 2026, according to Cryptopolitan reporting. The G1’s base price of $13,500 makes it one of the cheapest humanoid robots on the market, increasing the likelihood that these machines will appear at malls, schools, and public events.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

U.S. humanoid robot maker Agility Robotics is preparing to go public on Wall Street through a SPAC merger deal with Churchill Capital.
Agility announced Wednesday it has signed a definitive merger agreement with Churchill Capital, a blank check company backed by Wall Street dealmaker Michael Klein. The SPAC deal with Churchill Capital values the company at $2.5 billion pre-money.
Agility to trade under “AGLT” ticker
The combined entity will trade under the ticker “AGLT” on a major North American exchange.
Agility said the deal will generate more than $620 million in gross proceeds. Roughly $420 million would come from Churchill’s trust account, with a further $200 million raised through a common-stock PIPE priced at $10 per share, according to the announcement.
The $2.5 billion pre-money valuation represents a premium over Agility’s last private pricing. The company closed a $400 million Series C in March 2025 at a valuation of approximately $2.12 billion, and has raised more than $640 million across its funding history.
The SPAC deal remains subject to a Churchill shareholder vote and SEC review. However, both companies expect the transaction to close later in 2026.
Agility Robotics targets $1 trillion market opportunity
Agility, with its SPAC arrangement, is shooting for a market opportunity across manufacturing, distribution, and logistics environments in the U.S., which it estimates to be worth approximately $1 trillion.
The humanlike robot maker said it plans to use proceeds from the deal to fulfill existing customer orders, scale production of its Digit v5 robot, and expand commercial deployments, among other things.
Some of its competitors in the U.S. include Tesla, which makes Optimus, and Figure AI, which made the popular Figure 03 robot that walked side by side with the U.S. first lady, Melania Trump, during a summit at the White House, Cryptopolitan reported in March.
When completed, Agility will be the only publicly listed U.S. company focused entirely on humanoid robotics with active commercial deployments.
The company’s pitch to public-market investors rests on its paying customers, which most of its competitors lack.
Agility’s humanoid robot Digit currently operates in manufacturing, distribution, and logistics facilities run by Schaeffler, GXO, Toyota Motor Manufacturing Canada, and Mercado Libre, the companies disclosed.
Across nine customer sites, Digit has reportedly logged more than 65,000 hours of operation.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

The Reserve Bank of India (RBI) has released draft guidelines that would force banks and other regulated financial entities to adopt structures set in place to manage risks tied to artificial intelligence and machine-learning models.
These proposed risk management systems will include mandatory kill switches that instantly shut down any AI system producing harmful outputs.
The RBI’s interests
The proposed rules and guidelines to mitigate these AI risks drawn up by the RBI are open for public feedback until July 24. These guidelines cover every model used in bank operations, ranging from basic spreadsheet calculators to complex generative AI systems.
Banks would have to keep inventories of all models in their operation, assess risk at both the individual and enterprise level regularly, and take corrective action when risks cross acceptable thresholds. The modes of corrective action can range from tighter controls to complete decommissioning of the model, with well-detailed reports of the risks leading to such actions.
The framework also involves a risk classification system divided into multiple tiers. Every model, regardless of complexity, must be assigned a particular risk level, along with human oversight. High-risk models will require approval from the Risk Management Committee at the board level.
Kill-switches and human oversight for AI models
Banks are expected to build override, suspension, and deactivation mechanisms into every AI model deployment so that any model can be shut down immediately if it produces erroneous or harmful results.
The RBI also pointed to what it called an “automation bias,” which is the risk that bank employees will okay AI outputs without exercising any form of independent judgment. For customer-facing AI systems, the guidelines will mandate banks to let their customers know when they are interacting with an AI model, and give them the option to speak with a human at any point.
Generative AI models that interact with customers or external users will face additional cybersecurity requirements under the drafted proposal, according to Reuters.
Third-party AI providers come under scrutiny
Banks remain fully accountable for any AI model they choose to employ, regardless of if built internally or otherwise purchased from a vendor. The RBI also stated that supply chain concentration was one of its concerns, pointing to the risk that banks become overly dependent on a small number of global AI providers.
Independent validation is required for all third-party AI models. Banks must also define thresholds for AI and machine-learning models, ensuring they can articulate in plain terms why a model produced a given output.
The RBI has been tightening its technology governance position for a while. These draft guidelines represent the first time India’s central bank has proposed a comprehensive, board-level accountability structure specifically for AI model risks across the entire financial sector.
If you're reading this, you’re already ahead. Stay there with our newsletter.

A Chinese supercomputer has become the most powerful in the world, the first time since 2017 that a machine from China has topped a closely watched ranking often seen as a measure of a country’s technological strength.
The computer, called LineShine and based in Shenzhen, knocked the American machine El Capitan off the top spot in the latest TOP500 list released Tuesday.
Researchers behind the project said the computer, at China’s National Supercomputing Center, reached 2.198 exaflops, meaning it can run more than 2 quintillion calculations every second.
El Capitan, at the Lawrence Livermore National Laboratory in California, now sits second. LineShine runs entirely on regular processors rather than the graphics chips used for artificial intelligence, and it draws about 42.2 megawatts of power.
The milestone is one more sign of a wider shift
In a handful of industries China has pulled ahead of the United States, and few examples are clearer than a battery plant in the country’s southeast.
At a factory run by Contemporary Amperex Technology Company, or CATL, robot arms wind strips of metal into rolls and shape them into the bricks that become batteries. It is the largest and most advanced cluster of battery factories in the world.
For years the relationship ran the other way. American firms held the lead and came to China to build their goods cheaply, and Beijing made them share technology with local partners as the price of entry.
Now, in fields from batteries and solar panels to rare earths and life sciences, China is building some of the most advanced technology around and moving fast to corner new markets.
CATL shows the change. The company says it has a battery that can drive an electric car 250 miles on less than 10 minutes of charging, roughly three times faster than the batteries in most other electric vehicles. “Of course if there’s an opportunity in the U.S., we wish to pursue it,” Fred Zhang, a company spokesman told NYT.
The advances worry American officials, who see risk in leaning too much on Chinese technology. Representative John Moolenaar, a Michigan Republican who chairs the House Select Committee on China, said Beijing had subsidized CATL “to undercut non-China competitors and build worldwide dependence,” and called handing the company a critical industry “a grave error.”
Others warn that shutting out firms like CATL would leave American companies behind. “For decades now, we’ve been used to a world where the technology and innovation comes out of the West,” said Kyle Chan of the Brookings Institution. “The tables are turning.”
Automakers are split
General Motors has teamed up with South Korean battery makers, but CATL sells to Tesla and has licensed its technology to Ford for plants in Michigan and Kentucky. Ford ended a partnership with South Korea’s SK On in December. The company has drawn fire before.
In 2023, Virginia’s then-governor Glenn Youngkin blocked a Ford-CATL plant, calling it a “Trojan horse.” In 2025, Washington added CATL to a list of Chinese military companies, which the firm called “a mistake.”
Chinese investment faces steep hurdles in the United States, including federal security reviews, high tariffs, and a ban on Chinese vehicle software starting with 2027 models. Even so, companies are watching whether President Trump’s apparent openness to deals with China could open the door to more partnerships.
On May 14, Treasury Secretary Scott Bessent said a planned “board of investment” would decide which nonstrategic, nonsensitive areas might be open to Chinese money.
CATL was founded in 2011 out of a firm that made the battery for Apple’s iPod. It now produces about 40 percent of the world’s electric vehicle batteries and 30 percent of the batteries used to store solar and wind power. Of its 185,000 workers, 22,000 work in research, and more than 700 of them hold Ph.D.s.
A study released in June by the OECD found that Chinese industrial firms received three to eight times as much government support over two decades as companies in the 38 mostly wealthy nations in the group.
Pfizer’s chief executive, Albert Bourla, speaking in March, called China’s scientific “meteoric rise” a major challenge and predicted the country would pass the United States in drug innovation within this decade. “They built their science,” he said. “So this is where we need to become better.”
If you're reading this, you’re already ahead. Stay there with our newsletter.

Kalshi CEO Tarek Mansour has reportedly confirmed that the prediction market operator will not go public this year, even as the company’s annualized revenue has crossed $2 billion and its private valuation sits at $22 billion.
The company has reportedly held informal conversations with investment banks about a future listing. But people close to Kalshi’s finances say that a public offering would not come before late 2027 or early 2028 at the earliest.
What is Kalshi’s revenue performance since 2025?
Kalshi’s $2 billion annualized revenue figure is nearly a threefold increase from where the company stood in November 2025.
Sports prediction contracts have been the primary driver for this surge in revenue, with the NBA playoffs and the FIFA World Cup being responsible for most of the recent activities.
Monthly trading volume on the platform hit $16.81 billion in May 2026, a 13.5% increase over April.
During the opening week of the FIFA World Cup, Kalshi recorded $5.1 billion in weekly spot volume, which is the largest single-week total any individual prediction platform has recorded.
Institutional trading volume on the platform also grew 800% in the six months through early May, pushing annualized trading volume from $52 billion to $178 billion.
How much has Kalshi raised pre-IPO?
The IPO timeline discussion follows a fundraising streak for Kalshi, which closed a $1 billion round in May 2026. That round was led by Coatue, with Sequoia Capital, Andreessen Horowitz, IVP, Paradigm, Morgan Stanley, and ARK Invest among the participants.
The company has raised $2.685 billion across five rounds since June 2025.
State lawsuits continue to play spoiler
The regulatory drama around Kalshi and other prediction markets has become something of a weekly feature.
Just this week, Kentucky filed a suit against Kalshi and Polymarket, alleging that they are operating unlicensed sports betting platforms.
It joins Ohio, Nevada, New Jersey, New York, and Illinois, among others, to have made similar claims in court.
The CFTC has also been consistent with inserting itself into those proceedings, suing New Mexico after the state moved against Kalshi and arguing that federally regulated event contracts fall under its exclusive jurisdiction.
A New Jersey court sided with Kalshi in April 2026. However, the jurisdictional question is widely expected to reach the Supreme Court.
Gaming industry groups, including the American Gaming Association and the Indian Gaming Association, have lobbied Congress to add language to the CLARITY Act that would bar sports prediction markets from operating under federal derivatives rules.
Some estimates place sports contracts at up to 90% of Kalshi’s revenue, so the stakes are high for the platform.
Is the crypto IPO market cooling down?
Kalshi is not the only crypto-adjacent firm to postpone or shelve IPO plans in 2026. Many went into the year bullish on the possibility of listing, buoyed by the crypto-market performance in 2025 and the White House’s current pro-crypto stance.
However, the current market realities have made some firms have a rethink. Consensys, the company behind MetaMask, pushed its IPO to at least fall 2026.
Kraken has also suspended its own billion-dollar offering after confidentially filing with the SEC in late 2025. French hardware wallet maker Ledger also paused a planned $4 billion listing.
For now, BitGo remains the only crypto-native company to complete a U.S. IPO this year. As of mid-May, BitGo’s shares traded roughly 36% below the $18 offering price.
Bitcoin itself has lost about a third of its value in 2026, trading below $60,000 as of June 24, after hitting record highs above $122,000 late last year. Bitcoin ETFs have seen over $3.1 billion in net outflows this year, with more capital going toward AI and semiconductor stocks.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

A Hong Kong district court has rejected Xiao Rui’s claim that part of his money came from selling Bitcoin and found him guilty of laundering more than HK$64 million through underground banking channels.
The 37-year-old son of a former anti-corruption official from Wuhan faces sentencing on July 23 after being found guilty on four counts of money laundering and one count of using a false document.
Bogus BTC sale defense gets rejected
Between March 2014 and November 2023, about 38 transactions totaling over HK$64 million flowed into Xiao Rui’s personal bank accounts. The money came from at least 12 companies and 12 individuals. None had any business ties to Xiao’s asset management firm, yet he told the court that some of the money in his Hong Kong bank accounts came from selling Bitcoin.
Acting Judge Bernard Chung said Xiao could not provide any basic records like transaction dates, reference numbers, or wallet addresses to prove his claim. The judge found this explanation weak and rejected it outright.
Xiao also claimed that the money was a gift from his mother. According to him, she had earned the money through her business and gave it to him for investment in Hong Kong. However, under questioning, Xiao said his mother had worked as a nurse before moving to administrative jobs. He claimed she received a cash bonus of HK$20 million from a power company in 2016, but the judge said this story did not fit with her employment history.
The court heard that a contractor named Yao Qian, who had secured a municipal water-pump project in Wuhan, deposited about HK$4.72 million into Xiao’s bank account. Xiao Jun, Xiao Rui’s father, reportedly helped with this contract when he headed a bureau in the Wuhan People’s Procuratorate. Notably, the older Xiao was previously suspended amid a mainland corruption investigation.
The judge found the contractor’s testimony believable and rejected the defense claim that the money was legitimate business income.
How did Xiao get Hong Kong residency with fake documents?
The case also includes fraud related to Xiao Rui’s Hong Kong residency. In 2013, he applied to live in Hong Kong through the Capital Investment Entrant Scheme. This program required applicants to show they had assets over HK$10 million. Xiao submitted fake deposit certificates from a China Construction Bank branch in Wuhan.
Bank staff later confirmed the accounts did not exist. Despite this, Xiao was approved for residency in 2014, and that same year, he used his HSBC account to buy two fund products worth HK$10 million in order to meet the scheme’s investment requirements.
The court found Xiao guilty of using false documents for this application, and he claimed to be unaware of the fact that the documents were fake because his mother handled the application. The judge did not accept this excuse.
Prosecutors said the money laundering happened over nearly a decade. Xiao used accounts at Standard Chartered, DBS, and HSBC to move the money through what the court called underground banking channels.
The Independent Commission Against Corruption (ICAC) has asked for the criminal proceeds to be taken from Xiao, but a date for this hearing has not been set yet.
Xiao will stay in custody until his sentencing, where he faces prison time for the money laundering charges, which carry sentences of up to 14 years in Hong Kong.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Ortex Technologies, an analytics business, reports that short sellers are increasing their bets that Elon Musk’s SpaceX would continue to decrease after the company’s share price dropped from the highs it attained soon after going public on June 12.
The sale took place during a challenging period for the market as a whole.
The Nasdaq 100 was on track to lose more than $1 trillion in value on Tuesday as key IT businesses and semiconductor stocks declined.
SpaceX’s market value fell below $2 trillion for the first time since it began trading in the US.
Over the past three trading days, the business, which is anticipated to be included in the Nasdaq 100, has lost approximately $600 billion.
It would close with a market value of $1.95 trillion if Tuesday’s losses continued.
After peaking at $225.64 a few days after its launch, the stock is already down nearly 30%. Short sellers have been enticed into this downturn, which is part of a broader market decline, faster than many had predicted.
Data from S3 Partners indicates that roughly 40 million SpaceX shares are currently held in short positions, representing about 5%–7% of the company’s publicly tradable shares.
Recent figures suggest bearish bets are growing rapidly.
Ortex reports that short interest has climbed from 8% in the prior trading session to 13%, reflecting a sharp increase in the proportion of publicly available shares that have been sold short.
“A jump like this is a clear sign that a growing number of traders are positioning for the price to fall sharply,” stated Peter Hillerberg, co-founder of Ortex.
Borrowing the shares is getting easier
Traders are increasing their bets against the stock because it has become easier and cheaper to borrow shares for short selling.
“Shares are becoming more accessible,” said Sam Pierson, head of research at S3 Partners.
He noted that short sellers were paying about 0.60% annually to borrow shares.
While that is higher than the roughly 0.30% charged for the easiest-to-borrow stocks, it still suggests there is plenty of share supply available and fewer concerns about finding stock to short.
Ortex numbers, which demonstrated that the cost of borrowing is still low at roughly 1% and balances the supply of shares to lend against the desire to short a company, supported this.
At the start of trading, it hit 14%. There is still a lot of stock available for lending, according to Ortex, which says that utilization, or the percentage of available stock that is on loan, is currently at roughly 39%, up from the mid-30s last week.
The approach toward SpaceX is different from those of other large IT firms.
The Magnificent Seven, Apple, Microsoft, Nvidia, Alphabet, Amazon, Meta, and Tesla, only have 1% to 3% of their free floats traded short.
Ortex figures show that borrowing costs for those names are between 0.25% and 0.33%.
Options traders lean bearish
The derivatives market appears to be signaling a similar outlook.
According to Reuters, investors are assigning roughly a 40% probability that SpaceX’s share price will drop below $130 by mid-September.
Steve Sosnick, chief strategist at Interactive Brokers, noted that “The options activity has gotten more balanced.”
However, in several option series expiring between July and September, the number of outstanding put options, which gain value when the stock declines, is nearly double the number of open call options.
Ultimately, speculators who anticipate a decline will probably continue to be drawn to Musk’s rockets-and-AI company due to its high price tag.
Despite the $2 trillion value making it an apparent target, a number of factors, such as Musk’s history of openly opposing short selling and significant acquisitions by institutional and individual investors, may deter short sellers.
A request for comment was not immediately answered by SpaceX.
The aggressive shorting points to a change in how the market sees the company.
Investors appear to be looking past Musk’s long-term promises and focusing more on near-term spending, treating SpaceX as a heavy industrial business rather than a light software play.
The smartest crypto minds already read our newsletter. Want in? Join them.

Anthropic released a new Slack integration and named it “Claude Tag.” Enterprise and team users can now tag @Claude in Slack chats and delegate tasks.
The product replaces the existing Claude in the Slack app and represents a major expansion of how the company expects teams to interact with its AI. Claude Tag embeds a persistent AI into a team’s communication layer, which is better than one-off prompts or isolated coding sessions.
Anthropic’s Claude Tag works directly inside Slack
When an administrator connects Claude Tag to a Slack workspace, they get to define which channels, tools, and data sources Claude can have access to. Every configuration results in a unique scoped identity.
A Claude set up for a sales team operates with separate memory and permissions compared to another AI provisioned for engineering, according to Anthropic’s announcement. Administrators can cap token spend per organization or per channel and review a log of every action Claude has taken.
Claude Tag serves as a shared resource inside each Slack channel. One team member can assign a task, and another can pick up the thread later without having to re-explain the context. Claude breaks assigned work into stages, executes using connected tools, and posts results back into the Slack thread. Anthropic says 65% of its own product team’s code is now written by an internal version of the tool, and usage has spread beyond engineering into support, metrics analysis, and debugging.
Claude Tag is different from earlier Slack integrations. First, it accumulates context over time by following channel conversations, so users avoid repeating background information across sessions. Claude can also pull information from other Slack channels if granted read access, but does not surface content from private channels without permission.
Second, an optional “ambient” mode allows Claude to intervene without being tagged. When turned on, the AI pulls up relevant information from all connected or linked channels, follows up on stalled threads, and shows the team any changes it thinks they should see. Claude Tag can also schedule its own future tasks and work on multiday projects without continuous human input.
Anthropic moves deeper into enterprise turf
Claude Tag is part of a race among AI companies competing to become the relevant layer inside businesses.
Microsoft developed its Graph platform to supply organizational knowledge directly into Copilot. Snowflake and Databricks are branding their data platforms as backend repositories for AI agents. Glean is developing an intelligence layer that operates between models and enterprise data.
Anthropic took a different strategy by building the model within an existing communications medium. A huge chunk of internal business communication happens on Slack. This means that Claude Tag has access to institutional knowledge without adding a new integration layer.
The release adds to Anthropic’s growing portfolio of vertical products. Since February 2025, the company has shipped 13 AI tools geared toward specific use cases and industries. This includes agents for financial services, legal work, and small business operations, according to Cryptopolitan’s earlier reporting. The continuous stream of new products and features has created tension with partners who build on Anthropic’s models but now compete with Anthropic’s own products.
Claude Tag runs on Opus 4.8. Beta is available now to Claude Enterprise and Team customers, who can migrate from the existing Slack app within 30 days. The company offers introductory launch credits to eligible organizations.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

Bitcoin has dropped below $60,000 and is now down nearly 20% this month.
The total crypto market cap has fallen from $4.3 trillion in October to $2.0 trillion today.
Strategy shares hit their lowest level in more than two years as worries grow around its huge Bitcoin holdings.
Stocks are mixed as oil prices slide, Micron falls, and traders wait for the chipmaker’s earnings after the bell.

Leading prediction market platform, Kalshi, has filed a federal lawsuit against the state of Illinois, challenging the recently created law that would force prediction market platforms to obtain a state license.
The suit, filed in the U.S. District Court for the Northern District of Illinois, has named Governor JB Pritzker, Attorney General Kwame Raoul, and other state officials as defendants. Kalshi has also asked the court for a temporary restraining order and both preliminary and permanent injunctions to ensure the state of Illinois does not enforce the newly signed law on the intended July 1 date.
The new Illinois law
Governor Pritzker signed the SB3019 last week, a bill created as a state licensing requirement for prediction market operators and aiming to impose a 0.2% tax on all digital asset transactions involving customers based in Illinois, as previously reported by Cryptopolitan.
Kalshi, however, argues that these state-level requirements are invalidated by the Commodity Exchange Act because the platform operates as a CFTC-registered platform.
The prediction market platform’s argument further states that Federal law gives the CFTC exclusive jurisdiction over exchange-traded derivatives, and states cannot form additional regulatory regimes in addition.
Kalshi’s lawsuit
The lawsuit puts a spotlight on what Kalshi describes as an impossible choice between state and federal rules. If the platform pulls its sports event contracts from Illinois to comply with the new state law, that move would directly violate CFTC uniformity requirements.
Staying in the state without a license would also mean breaking Illinois law.
Implementing geoblocking technology to restrict Illinois users would be expensive and technically complex, Kalshi claimed in the submitted complaint. The company argues those costs would not be recoverable even if it ultimately prevails in its court lawsuit.
“It expressly violates the CEA’s ‘exclusive jurisdiction’ provision by asserting concurrent state jurisdiction over sports events contracts traded on federally regulated DCMs,” Kalshi said in its filing.
Federal vs State conflict ramping up?
The lawsuit is the latest in a growing number of disputes between the federal regulators at the CFTC and U.S. state governments over the control and regulation of prediction markets, most especially those offering bets tied to sporting events.
The Michael Selig-led CFTC has sued nine states so far to assert federal authority over the prediction market sector, including one filed against Illinois in April. The sued states have, however, pushed back against their federal counterparts, arguing that prediction market platform operations are essentially gambling products, and that they violate local betting laws.
If you're reading this, you’re already ahead. Stay there with our newsletter.

On June 24, humanoid robots kicked penalty shots at MWC Shanghai 2026 to try real-time AI decision-making. The 2026 FIFA World Cup is bringing football to the attention of people all over the world.
The Humanoid Robot Football Penalties Challenge took place at the Shanghai New International Expo Center in the Mobile AI Innovation Frontiers Zone. Robots must be able to read the goal, figure out angles, and hit without any help from a person or pre-programmed processes.
Models from Booster Robotics and Unitree Robotics were seen competing in videos circulating on the internet.
Robots score on perception, balance, and real time pressure
Each robot assessed the position of the ball and the goalkeeper’s movement before attempting a penalty kick. Real-time movement change is done based on data received from the robot’s sensors.
Organizers scored the robots on perception accuracy, balance control, motion planning, and adaptive response. The competition’s semi-finals wrapped on the first day and the final round is scheduled for June 25.
In a different story, Hyundai Motor started a football campaign with Boston Dynamics’ Atlas robot prior to the FIFA World Cup. Atlas watches football footage and tracks how players move and position their bodies. The robot then goes to a practice area to copy the movements it saw.
Sequences show Atlas shifting weight, swinging a leg to guide a ball with controlled contact, and progressing through balance and coordination drills. The robot raises its arms in celebration of a goal. In another scenario, it drops to one knee, mimicking an injury response from footage it had just watched.
Hyundai said that this is Atlas’s first time playing football as part of a “School of Football” program. There are plans to use Atlas and Spot, a four-legged robot, at the FIFA World Cup, but their tasks have not been finalized.
China’s humanoid advancement goes full speed
Penalty challenge fits a pattern. China has used high-profile public events to showcase its robotics sector, and this is another one. In April, Beijing held its second humanoid half-marathon, scaling from 21 robots in 2025 to more than 300 running alongside human competitors, Cryptopolitan reported. Organizers said 38% of participating teams entered an autonomous navigation group, where robots completed the course without human guidance. The Global Times characterized that shift as moving from “human-led mode” to “full autonomy.”
Counterpoint data cited by Cryptopolitan showed 16,000 humanoid robots were deployed globally in 2025, with China accounting for over 80% of installations. AGIBOT held the largest vendor share at 31.9%, followed by Unitree, UBTECH, and Leju. Tesla ranked fifth with about 5% share.
The smartest crypto minds already read our newsletter. Want in? Join them.

Top global crypto exchange, Binance, has officially withdrawn its MiCA license application from Greece’s Hellenic Capital Market Commission on Tuesday, stating that it will seek regulatory approval in a different EU member state after the process with the Greek regulator failed to reach a decision before the MiCA deadline.
The exchange announced this move in a post on its official X account, and has explained the reasons for the withdrawal in a blog post. The Binance announcement claimed that the exchange reviewed where the application to the Greek body stood and came to a conclusion that the current timeline no longer served its users.
CEO Richard Teng addressed European customers directly on X in a post which said, “We remain committed to securing a MiCA license in the coming months, while providing clarity, minimizing disruption, and keeping users informed directly.” He added that customer funds “remain safe and secure.”
To our European users: we understand regulatory uncertainty can be frustrating.
We remain committed to securing a MiCA licence in the coming months, while providing clarity, minimising disruption, and keeping users informed directly.
Your funds remain safe and secure. For… https://t.co/gIZUZ7KPhi
— Richard Teng (@_RichardTeng) June 24, 2026
Binance’s reasons for withdrawal
As part of the MiCA regulatory transition which commenced in late 2024, Binance submitted a licensing application to Greece’s Hellenic Capital Market Commission (HCMC). However, by the time the transition period expired, the exchange had neither received formal approval nor secured a license from the regulator.
Binance said in its blog post that the months of engagement with the HCMC were “constructive and in good faith,” however, the top exchange has chosen to redirect its efforts toward a jurisdiction that would give faster clarity on licensure requests. The company has not named the replacement EU country for the MiCA talks, stating that the new jurisdiction will be announced “when ready.”
Europe still a priority, Binance says
The exchange had reiterated that its pullback from Greece does not signal a retreat from the EU market in its entirety. Binance said it supports MiCA’s objective of building a unified regulatory framework and expects to obtain authorization from another member state “in the coming months,” according to both the blog post and Richard Teng’s X post.
The move leaves Binance without an active MiCA application in any publicly identified jurisdiction at a time when competitors including Coinbase, Kraken, and OKX have put in significant efforts to advance the acquisition of their own European licenses.
Are Binance users in Europe safe?
Some Binance users in Europe may face account-level changes depending on their country of residence and current account status, the company further explained. The exchange has said it will reach out to all affected customers through email and in-app notifications with details on required actions, the user’s options, and any relevant deadlines.
Binance has also extended a word of caution to users to ensure they verify communications carefully, warning that it will never initiate phone calls or request passwords, two-factor authentication codes, or private keys. CEO Richard Teng also reiterated in his X post that all user funds are “safe and secure.”
The smartest crypto minds already read our newsletter. Want in? Join them.

OpenAI will now use its first custom AI chip “Jalapeño” made by Broadcom. It will be used for the heavy computing work of running AI models for people using ChatGPT and other apps.
The company called it a big step in its plan to build the full stack behind its models and products.
“By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access,” Greg Brockman, OpenAI’s president, said in a statement.
Broadcom’s stock (NASDAQ: AVGO) rose about 2% after the news. The stock has grown 10% so far this year.
OpenAI helped start the rush into generative AI back in 2022, and since then it has been one of the largest buyers of Nvidia’s expensive graphics processing units, the main hardware used to build AI models and run big jobs. Since the demand of OpenAI’s products is too high, the company requires more sources for advanced inference chips.
For this reason OpenAI had also joined hands with Amazon Web Services for its Trainium AI chips. It has also looked for alternatives with Nvidia’s rival AMD and Cerebras which IPO’d recently as reported by Cryptopolitan.
After working together for 18 months, OpenAI and Broadcom had been in talks for a custom chip since October. They planned to build and roll out racks of OpenAI-designed chips starting with the goal of making enough to use 10 gigawatts of power.
What the new chip can do
The Broadcom chip is an ASIC. Industry experts say this type is less flexible than Nvidia’s GPU but costs less and can be built for certain AI jobs. OpenAI said it designed the chip in nine months and also built large parts of the computer system that will run it.
The two companies call the chip an “Intelligence Processor” and describe it as the first piece in a wider platform meant to make AI faster and easier to reach.
Broadcom has built a good reputation in the times of AI boom, with its assistance of custom chips for big cloud firms and leading labs.
A physical sample of the chip will reach OpenAI on Wednesday. The companies said they hope to start using Jalapeño chips by the end of 2026 and expand from there.
The move comes after months of tension over Nvidia hardware
Eight sources told Reuters that OpenAI has been unhappy with some of Nvidia’s newest chips and has looked for other options since last year. Seven sources said the company is concerned with how fast Nvidia’s hardware answers users on certain tasks, such as writing software code. OpenAI required chips would eventually handle about 10% of OpenAI’s inference needs.
However Altman said in a X post, “We love working with NVIDIA and they make the best AI chips in the world. We hope to be a gigantic customer for a very long time.”
Nvidia still leads in chips for training large models, but inference has become a new battleground. In September, Nvidia said it planned to invest as much as $100 billion in OpenAI, a deal that would give the chipmaker a stake in the startup.
In February, Nvidia CEO Jensen Huang also dismissed reports of friction as “nonsense” and repeated that Nvidia planned a large investment. Still, the $100 billion deal was scrapped and replaced with a much smaller one.
Later, Jensen Huang confirmed the full $100 billion was “probably not in the cards,” and he later described the original figure as “never a commitment”, just an invitation to invest up to that amount.
What Nvidia actually finalized was a $30 billion equity stake in OpenAI, committed at a $730 billion pre-money valuation. Huang said it was likely Nvidia’s last investment in OpenAI before the company’s planned IPO.
OpenAI’s new cyber model edges out Anthropic
OpenAI has been busy on another front too. On June 22, it fully launched GPT-5.5-Cyber as part of its Daybreak cyber defense program.
On CyberGym, a test built at UC Berkeley using 1,507 known software flaws from 188 open-source projects, the model scored 85.6%. Anthropic’s Mythos 5 scored 83.8% and Claude Opus 4.7 scored 73.1%.
Anthropic’s Mythos 5 and Fable 5 were pulled offline on June 12 after the Trump administration issued an emergency export control order, citing national security. As of June 23, both remain down with no set return date.
The smartest crypto minds already read our newsletter. Want in? Join them.

Sentient Foundation, a nonprofit aiming to support open, decentralized artificial general intelligence (AGI), announced a $42M grant program. The funds will be available to developers, researchers, and companies building open-source AGI tools.
Sentient Foundation offered $42M, a number famously known for the answer to life, the universe, and everything. The Foundation aims to democratize the process of building AGI tools and reach a wider audience with the benefits of powerful models.
The Foundation already noted a trend of improving open-source AGI projects, expecting some of the new models to rival Anthropic’s Fable 5.
Sentient Foundation believes AGI development will accelerate, hence the significant commitment to advancing open-source AGI building. The program will offer non-dilutive grants and founder-friendly investments, aiming to accelerate the growth of AI infrastructure, models, tools, and applications.
The grant program will be open to global companies for building, innovating, or scaling existing models. The applications are open immediately and will be evaluated on an ongoing basis.
Sentient Foundation secures open-source funding mechanism
AI and AGI development within the confines of leading companies leaves out the open-source space. Sentient Foundation aims to build a clear and dedicated funding mechanism for this space.
“The future of intelligence should be built by the many, not controlled by the few,” said Sachi Kamiya, Director of Venture and Growth at the Sentient Foundation.
“A few companies are trying to become the OPEC of intelligence — meter it, price it, decide who gets it. We’re making it air,” she said.
Grants from Sentient Foundation can be tailored to projects at any stage of development. The funding can go toward researchers, open-source maintainers, independent developers, and public-goods initiatives. There will be no specific equity requirements, ownership claims, or restrictions on future development. Grant recipients will retain ownership of their work and intellectual property.
For teams developing commercial businesses around open-source AGI, Sentient Foundation will offer an Investment Track approach. Those investments will be founder-friendly, allowing teams to scale while retaining the open-source access.
Applicants will be screened based on technical merit, ecosystem impact, openness, and long-term potential. The Foundation has built an advisory council of experts from the open-source AI community to judge projects.
Projects may keep some of their components as closed-source, but must make at least one essential element of their project openly available and meaningfully tied to the project’s value and adoption.
Sentient Foundation aims to build on existing open-source AI projects
Sentient Foundation pointed out that AI development had multiple successful AI initiatives with open source. Projects like Ollama, Llama.cpp, LeRobot, DeepSeek, and others have shown technologies ready for global adoption.
Sentient has already supported open projects like ROMA, Open Deep Search, EvoSkills, Arena, and its own OML family of models.
“Open models are improving at an extraordinary pace,” Kamiya said. “When they catch up, and they will, the people building on them should win, not pay rent on intelligence forever.”
Sentient Foundation will run its initiative with support from other ecosystem participants, including Alibaba Cloud, Franklin Templeton, and academic institutions like Princeton University and the Indian Institute of Science.
The smartest crypto minds already read our newsletter. Want in? Join them.

The Commodity Futures Trading Commission (CFTC) filed a one-count complaint against Kentucky on Tuesday, June 23, to block the state’s enforcement actions against Kalshi and Polymarket. The lawsuit names the Commonwealth of Kentucky, Governor Andy Beshear, Attorney General Russell Coleman, Department of Revenue Commissioner Thomas Miller, and the Kentucky Racing and Gaming Corporation as defendants. It asks the court to declare Kentucky’s prediction-market laws unconstitutional and to bar their enforcement.
The complaint calls Kentucky’s actions “the latest entries in Kentucky’s campaign to banish prediction markets from within their borders.” Kentucky becomes the ninth state the CFTC has sued in its campaign to assert exclusive jurisdiction over event contracts, after Illinois, Arizona, Connecticut, New York, Rhode Island, Wisconsin, Minnesota, and New Mexico.
Prior targets included governors and attorneys general who were Democrats. In contrast, Kentucky does not follow this trend. While Beshear is a Democrat, Coleman is a Republican attorney general, which makes him the first such attorney general to face a CFTC suit.
Coleman’s June 17 suit triggered the federal response in six days
Coleman filed separate lawsuits against Kalshi and Polymarket on June 17 in Franklin Circuit Court, accusing both platforms of running unlicensed sportsbooks under Kentucky gambling law. “Kalshi and Polymarket are operating illegal sportsbooks in Kentucky and breaking our laws,” Coleman said in the announcement.
The Kalshi suit also named Coinbase, Robinhood, and Webull as affiliated defendants, alleging the platforms helped Kentucky users access sports event contracts. The Polymarket suit alleged the platform spread false and misleading advertising suggesting it could legally offer sports betting in Kentucky.
Kentucky’s complaint claimed 89% of Kalshi’s $23 billion in 2025 contract volume came from sports wagering. Both Kalshi and Polymarket have already removed the Kentucky cases from state court to federal court, betting on a more favorable venue.
Kentucky may attempt to push the cases back to state court, which would set up the procedural fight Coleman has signaled he wants.
The Michigan ruling last week gives the state side new leverage
A federal judge in Michigan’s Western District denied Polymarket’s request for a preliminary injunction against state regulators, ruling that sports prediction markets do not qualify as swaps and therefore fall outside CFTC exclusive jurisdiction.
Polymarket has appealed to the Sixth Circuit, which also covers Kentucky, Ohio, and Tennessee. Two district court judges in the Sixth Circuit have now preliminarily sided with state regulators while one has sided with the platforms.
The Third Circuit sided with Kalshi in the New Jersey case earlier this year. A federal judge paused Arizona’s criminal case against Kalshi after the federal preemption argument was deemed likely to win.
The CFTC’s litigation strategy is partly designed to push the conflict toward a Supreme Court ruling that legal observers expect within 12 to 18 months.
CFTC Chairman Michael Selig said in April that the agency would defer to whatever the courts decide. That position now meets a Sixth Circuit appeal where the federal jurisdiction question is the actual issue.
The 14.25% Kentucky tax is the second front
Kentucky also enacted a 14.25% excise tax on prediction-market transaction fees in April under House Bill 904, set to take effect January 1, 2027. It would be the first such tax in the country. The Coalition for Fair Markets, a trade group representing Kalshi, Crypto.com, and Polymarket, filed its own lawsuit against the Kentucky attorney general on June 12, calling the tax discriminatory against federally regulated platforms.
The CFTC’s June 23 complaint addresses the tax directly, arguing it “concerns or regards” exchange-traded derivatives within the agency’s exclusive jurisdiction under the Commodity Exchange Act.
The combined federal pressure (the CFTC suit against the AG and the industry coalition suit against the tax) puts Kentucky in the same defensive position as Wisconsin and Minnesota, but with the partisan optics flipped.
More than a dozen other states have sent cease-and-desist letters or taken legal action against prediction-market operators, including Montana, Nevada, Utah, Iowa, Ohio, Tennessee, New Jersey, Maryland, and Connecticut. 41 state attorneys general urged Selig to affirm state authority over gambling, and 39 AGs filed an amicus brief backing Massachusetts in its case against Kalshi.
The Kentucky suit weakens the partisan-cover critique of the CFTC’s campaign
A New York Times investigation in May reported that senior career CFTC officials were suspended or pushed out after raising regulatory concerns about Polymarket, Crypto.com, and a Gemini affiliate, all firms with documented Trump family financial ties. Donald Trump Jr. is a paid adviser to Kalshi and a Polymarket investor through 1789 Capital. President Trump backed the CFTC’s exclusive authority over prediction markets in a Truth Social post earlier in June.
As Cryptopolitan reported on June 19, Kalshi crossed $2 billion in annualized revenue at a $22 billion valuation with monthly trading volume of $16.81 billion in May. If state AGs are correct that 89% of contract volume came from sports, an adverse ruling could eliminate the primary revenue source the company is preparing to take public in 2027 or 2028.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Adrian Boafo has won the Democratic primary for Maryland’s 5th Congressional District, securing a major victory for the cryptocurrency industry’s political ambitions.
The win was encouraged by roughly $5.5 million in spending from Protect Progress, the Democratic arm of crypto super PAC network Fairshake.
How much did crypto PACs spend on Boafo?
Adrian Boafo has won in the crowded 24-person Democratic primary for Maryland’s 5th Congressional District. Former Capitol Police officer Harry Dunn and Quincy Bareebe ran alongside Boafo.
Protect Progress, the Democratic arm of the crypto super PAC network Fairshake, spent approximately $5.5 million to propel Boafo’s win, according to the Associated Press and Federal Election Commission filings.
The total outside spending on Boafo’s behalf reached roughly $11 million when including support from pro-Israel groups. The United Democracy Project, a super PAC tied to the American Israel Public Affairs Committee (AIPAC), contributed $5.7 million to the race, largely to support Boafo.
Boafo’s own campaign raised $1.14 million while Oracle executives, including Executive Vice President Kenneth Glueck and SVP of Government Affairs Josh Pitcock, accounted for $26,500 in direct contributions. Coinbase employees, led by VP of Engineering Jesse Pollak, added $11,500.
Protect Progress also directed smaller sums to oppose Dunn ($172,295) and Bareebe ($98,454), according to Tech Influence Watch, a spending tracker run by independent journalist Molly White.
During the 2024 election cycle, the Fairshake network spent $133 million backing crypto-friendly candidates in both parties. For 2026, the network has raised between $150 million and $200 million from contributors, including Coinbase and Ripple.
Protect Progress has already helped tip other Democratic primaries this cycle. For instance, the PAC spent more than $600,000 supporting Melissa Bean’s winning House campaign in Illinois. Last month, pro-crypto Rep. Christian Menefee won against Rep. Al Green in a Texas Democratic primary runoff after receiving crypto super PAC endorsements.
Each primary victory adds another vote sympathetic to pending legislation like the CLARITY Act, which would draw jurisdictional lines between the SEC and CFTC over digital assets.
However, the same night Boafo won, Rep. Dan Goldman lost the Democratic primary in New York to Brad Lander. Goldman had voted in favor of both the GENIUS Act and the Clarity Act when the House passed the two crypto-related bills last year.
Why did Boafo get millions from crypto PACs?
Boafo’s background and policy positions make him a natural ally for the digital asset industry. He served as Oracle’s director of government affairs starting in March 2021, while simultaneously sitting in the Maryland House of Delegates after taking office in January 2023.
Oracle is a cloud infrastructure provider whose services are used by crypto exchanges and decentralized finance projects.
As a state legislator, Boafo established a strong pro-crypto record. He sponsored House Bill 470, which created a task force to study and make recommendations on the use and regulation of blockchain technology and digital assets in Maryland.
In a 2025 LinkedIn post promoting the legislation, Boafo referred to blockchain as “the future.” Gov. Wes Moore signed the bill into law earlier this year in May.
Boafo also sponsored the Maryland Financial Innovation Act of 2026, which aims to establish clearer regulatory frameworks for digital assets.
The smartest crypto minds already read our newsletter. Want in? Join them.

DeFi protocols have shed roughly $45 billion in total value locked across all chains since January, amid a record wave of exploits and price crashes.
On-chain analysis by CryptoRank showed that DeFi TVL has been falling every single month this year, sliding from about $115 billion to around $70 billion as of late June.
Among the ten largest blockchains by TVL, only Tron and Hyperliquid posted positive growth this year.
DeFi TVL year-to-date. Source: CryptoRank
TRON gained about 5%, largely due to its dominant role in USDT transfers and stablecoin settlement. Hyperliquid rose roughly 7%, driven by its position as the leading venue for on-chain perpetual futures trading.
Arbitrum took the steepest hit among top-ten chains, dropping 55.3% to $1.3 billion in TVL.
Ethereum has seen a 43% year-to-date decline but remains the largest DeFi chain, with $38.9 billion in TVL. Solana fell 40.5% to $4.93 billion.
DeFi TVL decline amid crypto market crash
The outflows closely follow the broader crypto price corrections, which began after Bitcoin hit an all-time high above $122,000 in October 2025, with the total crypto market capitalization peaking at $4.21 trillion.
At the time of writing, the crypto market sits at a $2.15 trillion capitalization, which represents nearly 50% drop from the October peak.
BTC is down over 28% year-to-date. ETH sees as high as 43% decline, followed by BNB and SOL at 33% and 43.5% respectively, which account for the second and third-largest chains by TVL.
DeFi hacks compound pressure
CryptoRank also attributed the outflows to a record wave of exploits that have hit the market this year.
The crypto industry has recorded 121 hacks in 2026 through late June, with combined losses reaching approximately $942 million. The second quarter alone accounted for 85 of those incidents and roughly $775 million in stolen funds.
In Q2, KelpDAO lost $293 million through a LayerZero cross-chain bridge vulnerability, and Drift Protocol suffered a $280 million breach, Cryptopolitan reported.
The KelpDAO attack also had an impact on Aave. The attackers used stolen, unbacked rsETH tokens as collateral on Aave to borrow against them, leaving the lending protocol with bad debt.
In the days following the exploit, Aave’s TVL dropped from $26.4 billion to $14.3 billion, a 46% decline, as depositors pulled funds.
The KelpDAO and Drift Protocol exploits alone accounted for three-quarters of Q2’s total losses.
If you're reading this, you’re already ahead. Stay there with our newsletter.