Senate Democrats called for hearings into President Trump’s crypto holdings after disclosures showed over $1.2 billion in crypto-related income.
The lawmakers cited potential conflicts of interest, foreign influence, and Trump’s role in shaping crypto policy.
The push comes as the Clarity Act remains stalled over ethics provisions targeting Trump’s crypto ventures.
Democratic senators are calling for congressional hearings into President Donald Trump’s crypto holdings after new financial disclosures revealed they generated over $1.2 billion in income for him last year.
Five Senate Democrats—Elizabeth Warren (D-MA), Richard Blumenthal (D-CT), Gary Peters (D-MI), Dick Durbin (D-IL), and Ron Wyden (D-OR)—said the disclosures raised concerns about potential conflicts of interest, foreign influence, and Trump’s role in shaping crypto policy while profiting off the industry.
“The disclosures heighten concerns about the president pushing Congress to pass crypto legislation in favor of the very industry he’s cashing in on, the administration’s moves to exempt cryptocurrencies and service providers from existing financial services regulations, and its steps to weaken enforcement, including by disbanding the Department of Justice’s National Cryptocurrency Enforcement Team,” the lawmakers said in a statement Friday.
The senators who signed today’s letter are the top Democrats on the Senate Banking, Investigations, Homeland Security, Judiciary, and Finance committees, respectively. They have requested that all of those bodies hold hearings on the president’s crypto dealings.
The group specifically underscored how Trump’s disclosures showed that unknown “third-parties” own a stake in his family’s crypto company, World Liberty Financial. UAE royals purchased a 49% stake in the firm last year.
The request for hearings comes as the Trump administration pushes Congress to pass the Clarity Act, a landmark bill that would formally legalize most crypto activity in the United States. The legislation has stalled over numerous issues, chief among them disagreements over ethics provisions targeting the president’s ability to issue and endorse digital assets while in office.
Trump’s financial disclosure report, released last month, showed he earned more than $1.2 billion from crypto-related ventures, including more than $635 million tied to his Trump meme coin and over $588 million from token sales connected to World Liberty Financial. The disclosures also showed Trump holds tens of millions of dollars worth of Bitcoin and Ethereum.
The Clarity Act advanced out of the Senate Banking Committee in May after two Democrats split from their party to support moving the bill forward—though the lawmakers warned a deal on ethics language remained necessary to earn their votes on the Senate floor.
The legislation’s supporters argue it needs to clear Congress by August to become law this year, given the looming November midterm elections.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
Strategy has started selling Bitcoin to fund dividends on its preferred stock, a shift from its long-standing “never sell” stance that has unsettled the market.
Standard Chartered calls the selling “mostly noise” and a communication problem, and is sticking to its end-2026 Bitcoin forecast of $100,000.
Bitcoin trades around $64,440, up 3.8% on the week but down 42% on the year, while Myriad traders give Strategy a slim chance of holding 1 million BTC this year.
Strategy’s turn from hoarding Bitcoin to selling it, in order to pay dividends on its preferred stock, has “muddied” Bitcoin’s near-term prospects, according to a new Standard Chartered note that nonetheless urges investors to look past it.
Last week, Bitcoin treasury company Strategy sold 3,588 BTC for about $216 million between June 29 and July 5 to cover preferred-share dividends and top up a cash reserve, leaving it with 843,775 BTC. That came after a token sale of just 32 BTC in early June that triggered its worst week since 2022.
In a note published Friday, Standard Chartered’s Geoff Kendrick wrote that the bank sees the sales as “mostly noise rather than a signal” of Bitcoin’s medium-term direction, and kept its end-2026 forecast of $100,000.
Strategy’s mNAV issue
For years Strategy ran a simple machine: as long as its shares traded well above the value of its Bitcoin holdings, a premium captured by a metric called mNAV, it could issue stock, buy more BTC, and lift both its own valuation and Bitcoin’s price. That premium has evaporated.
Standard Chartered puts mNAV at around 1 on an enterprise-value basis, while equity-based trackers land lower: BitcoinTreasuries has the stock at around 0.7 times the value of its Bitcoin on a diluted basis, a discount of roughly a third. Either way, the machine has stalled. Strategy’s BTC stack, bought for $63.7 billion, is worth around $54 billion at current prices, and the firm booked an $8.3 billion loss on its digital assets last quarter, almost all of it unrealized.
Backing the STRC dividend
With the accumulation model stalled, Strategy is repurposing its Bitcoin as collateral for STRC, a perpetual preferred stock known as “Stretch” that pays a 12% annual dividend and has about $10 billion outstanding, per Standard Chartered.
The shares are designed to trade near their $100 par value, but they slid to an intraday low of $71.25 on June 26 after the company disclosed its first Bitcoin sale earlier that month. Price action since suggests “the market has yet to be fully convinced of this pivot,” Kendrick wrote. Under a “BTC Monetization Program” unveiled on June 29, Strategy can raise up to $1.25 billion by selling Bitcoin to keep the dividends funded.
Kendrick argued that clear communication is “key to reassuring markets that wholesale selling is unlikely,” which should pull STRC back toward its $100 par and, in turn, ease pressure on Bitcoin. Because the stock is “heavily over-collateralised” by the Bitcoin behind it, he wrote, effective signaling could remove the need for Strategy to sell any more at all. The analyst noted that the reserve behind the dividend now holds $2.55 billion, almost a year and a half of coverage.
What the market thinks
Bitcoin changed hands around $64,440 on Friday, up 3.8% on the week but down 42% over the past year and about 49% below its October 2025 record of $126,080, per CoinGecko data.
Traders are doubtful Strategy’s buying spree will resume at full tilt. A market on Myriad, a prediction platform owned by Decrypt’s parent company Dastan, puts the chance of the company holding more than 1 million BTC before 2027 at around 13%. Strategy holds 843,775 today, more than 4% of all the Bitcoin that will ever exist.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
Researchers introduced “Adversarial HalluSquatting,” an attack that exploits AI-generated hallucinations.
The technique tricks AI agents into trusting fake repositories or tools that contain malicious instructions.
Tests against popular AI coding assistants showed the method could lead to remote code execution in controlled experiments.
AI hallucinations may be more than incorrect answers—they could become a way for hackers to compromise computers, according to new research from Tel Aviv University, Technion, and Intuit.
In the paper, “Beware of Agentic Botnets: Scalable Untargeted Promptware Attacks via Universal and Transferable Adversarial HalluSquatting,” researchers demonstrated a technique that exploits AI models when they generate fake links to software repositories and other online resources.
“The growing adoption of agentic LLM applications has introduced a new threat previously named as promptware,” the researchers wrote. “While prior work has established that adversaries can exploit direct channels to LLM applications to apply promptware under weak threat models, many applications do not provide any direct channels that could be exploited for prompt injection beyond the Internet.”
Known as adversarial hallucination squatting or “HalluSquatting,” the attack involves predicting which fake resources AI models are likely to create, registering those names, and adding malicious instructions. If an AI agent later retrieves the hallucinated resource, it may treat the attacker-controlled content as legitimate.
The researchers said the threat emerges as AI assistants move beyond answering questions and gain the ability to interact with computers—accessing files, searching the web, writing code, and running commands.
Those abilities can create security gaps when agents act on information they retrieve without confirming whether the source is real.
“Ongoing studies have demonstrated various variants of Promptware attacks against real-world systems, including ChatGPT, Google Assistant, Copilot, and various additional applications,” they wrote. “These works demonstrated that Promptware can lead to financial, privacy, and safety impacts.”
Researchers warned the technique could allow attackers to build AI-enabled botnets. A botnet refers to a network of infected computers or devices controlled remotely by an attacker. Botnets are commonly used in cyberattacks, including denial-of-service attacks, cryptocurrency mining, malware distribution, and ransomware campaigns.
In testing, the researchers found AI-generated resource hallucinations occurred at rates as high as 85% in repository cloning scenarios and 100% in skill installation tests.
The team evaluated the technique against AI coding assistants and agents, including Cursor, GitHub Copilot, Gemini CLI, and OpenClaw.
HalluSquatting is similar to typosquatting, a cyberattack tactic where attackers register domain names resembling legitimate websites or software packages to trick users. Instead of exploiting human typing mistakes, HalluSquatting targets mistakes made by AI models.
The news comes as researchers continue to test how attackers can manipulate AI agents.
In April, Google researchers detailed malicious websites designed to hijack AI agents through indirect prompt injection attacks, including attempts to steal passwords, delete files, and manipulate payments. A separate study on the “CopyPasta” attack showed how hidden prompts inside developer files could manipulate AI coding assistants into spreading malicious code.
In June, an OpenClaw user reported facing more than 6,000 attempts from attackers attempting to trick the AI agent into leaking sensitive information.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
On Thursday, MoonPay launched support for MoonAgents on Telegram, bringing its AI crypto assistant to the messaging platform.
Users can ask MoonAgents to analyze markets, create dashboards, and prepare transactions using natural conversations.
MoonPay says Telegram acts as an interface, while user data and private keys remain on their computers.
MoonPay is bringing its AI crypto assistant to Telegram, letting users interact with MoonAgents through the messaging platform, the company announced on Thursday.
According to MoonPay Agents Product Lead, Kevin Arifin, the integration builds on the company’s MoonAgents desktop app by giving users access to their AI assistant when they are away from their computer.
“The desktop app is tied to your computer, but sometimes you want to make trades on the go or do analysis when you’re out for a walk,” Arifin told Decrypt in an interview. “The Telegram integration is the gateway to do that.”
Similar to OpenClaw and Hermes Agent, MoonAgents uses Telegram as an interface for users to interact with an AI agent. Arifin said users create a custom bot through BotFather, connect it to the MoonAgents desktop app, and can ask the agent to analyze markets, prepare transactions, and monitor blockchain activity. (Disclosure: MoonPay Ventures is an investor in Dastan, Decrypt’s parent company.)
While much of the cryptocurrency space utilizes Telegram to communicate, Arifin said Telegram was chosen not only because of its popularity among crypto users.
“The reason we chose Telegram as the first integration for the Moon Agents desktop app isn’t really that all of crypto uses it,” Arifin said. “They provide a really great interface for creating a new bot,” he said, noting that other messaging platforms introduce more friction, pointing to the additional setup required for similar integrations with services like WhatsApp or iMessage.
According to Arifin, AI agent projects like OpenClaw and Hermes Agent also influenced MoonAgents by showing how AI assistants can operate outside traditional chatbot interfaces.
“I think OpenClaw really redefined what the experience for LLMs could look like, especially with LLMs that can access your computer,” Arifin said.
While Telegram provides the interface, Arifin said MoonAgents was designed so users are not dependent on the messaging platform to access their agent.
“I think the greatest part about this is the focus of the Moon Agents Desktop App is everything is saved on your computer, so even if Telegram shuts down tomorrow and you can’t take your agent on the go anymore, all your conversations still exist on your computer,” Arifin said. “You can continue to have a conversation with your agent through the Moon Agent Desktop App, and essentially continue that conversation, just like Telegram never existed.”
The news comes as crypto companies continue developing infrastructure that lets AI agents interact with digital assets and online services. In April, Gemini launched Agentic Trading, which allows users to connect AI models, including ChatGPT and Claude, to execute trading strategies through the exchange’s tools.
In June, Coinbase launched Coinbase for Agents, a tool that allows AI agents to trade crypto, make payments, and manage portfolios within user-defined limits. Also in June, Nous Research released a desktop version of Hermes Agent, moving the open-source AI agent from a command-line tool to a standalone app.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
Bitcoin fell 2.89% this week, closing at $61,749 after failing to break resistance in the $64–65K range—the key zone bulls needed to reclaim to change the short-term narrative.
Ethereum confirmed a weekly death cross for the first time in years, with its 50-week EMA now below its 200-week EMA, and prediction market traders now pricing a 72.3% chance ETH hits $1,500 before it sees $3,000 again.
The broader crypto Fear & Greed Index sits at 23 (extreme fear), spot Bitcoin ETFs just ended a 10-day, $2.7 billion outflow streak.
The crypto market enters the second week of July in rough shape.
Bitcoin is holding on, but just barely, in the low $60,000s after briefly touching 21-month lows under $58,000 last week. Ethereum is below $1,750, down around 4% on the day, and more than 30% in the last year. The broader market is down, of course, and altcoins are down harder.
The total crypto market cap excluding BTC and ETH shed 30% since January. Crypto IPOs—Gemini, Bullish, BitGo—have imploded since their debut.
The mood is, understandably, grim.
But grim moods have a long history of being wrong at exactly the wrong time. Every major Bitcoin bear cycle since 2009 has ended with a flush, an extreme fear reading, and a moment where the obvious trade looked like going short.
Bitcoin has now been through four such cycles, and in nearly every case, a pre-halving compression phase—where price grinds lower and sentiment deteriorates before the next supply shock—preceded the next leg up. The next halving—when mining rewards, and therefore the supply of newly minted Bitcoin, are cut by 50%—is roughly 21 months away, which historically is when accumulation starts making uncomfortable sense.
The difference this cycle? Crypto is now mainstream.
Spot Bitcoin ETFs, institutional balance sheets, formal accounting standards changes, and a legislative framework for digital assets have all arrived since the last halving. Bitcoin now has a fundamentally different institutional status than it did when BTC was a niche hobby. That doesn’t eliminate volatility—it just means the players in this bear market are wearing different suits than last time. Whether that speeds up or delays the bottom is an open question. The charts, for now, have their answer.
Bitcoin price: optimism with an asterisk
Bitcoin opened the week at $63,587, hit a high of $64,657, then closed lower, meaning that the bulls showed up, tried to push through, and failed. Bitcoin is trading hands at $61,749, down 2.89% in the week.
It’s important to note that BTC fell to $58,035 just days ago—a 21-month low—before bouncing.
The resistance zone that stopped the spike is exactly the one everyone was watching. The $64–65K area has been acting as a ceiling since early June, and this week’s candle barely kissed it before retreating. On Myriad, a prediction market developed by Decrypt’s parent company Dastan, traders are placing nearly 73% odds that Bitcoin touches $55,000 before $84,000. The sentiment among predictors flipped on June 2—before that, the smart money was leaning bullish.
Zooming out on the weekly chart, the Fibonacci retracement (natural support and resistance zones that happen during a trend) of that entire downleg from $82,833 places the $73,245 and $70,284 zone as with the most activity.
The Average Directional Index, or ADX, is at 30.7. The ADX measures trend strength regardless of direction on scale from 0 to 100. When it’s above 25, this tells traders that an actual trend is in place, and 30.7 is solidly there. Based on directionality, bears are in control.
The Relative Strength Index, or RSI, sits at 36.8. RSI measures momentum, similarly on a 0–100 scale: Above 70 signals overbought conditions and usually triggers profit-taking; below 30 signals oversold conditions that typically attract buyers. At 36.8, Bitcoin is close to oversold but hasn’t crossed the threshold yet. The technical setup suggests selling pressure may be approaching exhaustion—but “approaching” isn’t “done.” Right now markets appear to be panic selling.
One note of caution for the bears: The picture painted by the exponential moving averages remains bullish. Bitcoin’s 50-week exponential moving average, or EMA, is still above its 200-week EMA. When this happens, it forms a pattern that traders refer to as a “golden cross,” which in this case is technically still intact. But it’s narrowing fast. The inverse of a golden cross is a death cross, and if it forms on the weekly chart it would represent a structural shift that very few Bitcoin cycles have survived without a deeper flush first.
Thankfully for permabulls, this has not happened in a while.
Reasons for the bullish case are mostly fundamental:
Spot Bitcoin ETFs just snapped a 10-day, $2.7 billion outflow streak with a $221.7 million single-day inflow on July 2, and have since pulled in roughly $510 million. On-chain data from Glassnode shows long-term holders have returned to accumulation after an extended period of distribution, with buying activity broadening across wallet cohorts.
The Fear & Greed Index at 23, registering “extreme fear,” is historically a contrarian signal—not a guarantee, but a pattern. Some indicators approaching oversold from the weekly chart suggest the selling may be closer to exhausted than just starting.
$BTC has seen a series of bullish patterns broken, evidence of the power of the downtrend. Will this ‘W’ be the one that breaks the trend?
— John Bollinger (@bbands) July 2, 2026
For the bearish scenario, the technicals are more apparent for those focusing on shorter time frames:
Bitcoin failed to break the exact resistance everyone was watching. ADX at 30.7 with bearish directional index confirms an active downtrend with real momentum. Year-to-date ETF outflows are still negative. Citi downgraded its 12-month Bitcoin forecast to $82,000 with a bear case at $53,000. The Fibonacci target below current price at $57,735 is still the most visible technical magnet on the chart. Myriad’s prediction market—where money, not opinions, speaks—says 72.3% chance of $55K first.
Ethereum price: The death cross nobody wanted
Ethereum is trading at $1,729.7, down 3.06% from its $1,784 weekly open. That number is painful enough. But the bigger story isn’t the weekly candle—it’s what just happened on the weekly chart under the hood.
Ethereum has just confirmed a weekly death cross. The 50-week exponential moving average has crossed below the 200-week EMA for the first time in years. The upcoming days/weeks will be key to define positions for long-term trades if the cross extends and is not invalidated.
On shorter timeframes, death crosses happen regularly and can reverse quickly. On the weekly chart, they represent months of structural deterioration, and they tend to define entire market phases rather than single moves.
Ethereum’s daily chart has been in death cross since November 2025, when ETH peaked near $4,100 before beginning its extended decline. That daily bearish structure has now propagated to the weekly frame—a longer-timeframe confirmation that the bear trend isn’t a blip.
Traders on Myriad appear as bearish on ETH as they do on BTC, likewise pricing in a 72% chance Ethereum hits $1,500 before $3,000. These odds flipped in May—before that, the market was closer to 50-50 between the two outcomes. The gap between options is now at its largest since June, suggesting conviction has moved firmly into the bearish camp among traders putting actual money on the line.
The Fibonacci retracement on ETH’s downleg from $2,465.8 to $1,505.1 defines the zone between $2,098.9 and $1,985.5 as the ones with the most activity to watch for. Current price at $1,729.7 is pinned near the Fib level at $1,731.8. Below that, the next meaningful technical reference is the $1,500 price zone. That’s exactly the doom scenario Myriad traders are betting on.
The ADX reads 26.5 with bearish directionality—same story as Bitcoin, just more pronounced. A trend is confirmed, the direction is down, and the bears have the momentum. RSI at 36.9 mirrors Bitcoin’s reading almost exactly: bearish, approaching oversold but not there yet.
Some hopium for the bulls: Weekly death crosses on Ethereum have historically appeared around the final stages of bear market cycles—not the middle of them. In prior cycles, the three-day death cross frequently coincided with or immediately preceded significant bottoms. In other words, this is the panic zone in which many people wait to buy the asset for cheap.
If that pattern holds, the pain may be closer to ending than beginning. ETH spot ETFs turned positive on July 2 with $29.1 million in inflows. RSI is approaching oversold on the weekly—a zone that has historically been a strong accumulation signal for patient buyers.
Now for the bears: A weekly death cross is a new structural reality, not a temporary signal—it took months to form and typically takes months to reverse. US spot ETH ETFs logged a record 17 consecutive days of net outflows totaling $401 million in May, followed by another 10-day streak in June.
The Fibonacci target of $1,500 is technically the next major level, and it’s the exact number Myriad’s 72.3% majority is betting on. Citi’s bear case for ETH is $1,094. The weekly structure doesn’t give bulls much to work with until the price of Ethereum reclaims the $2,000 area—a 15.6% climb from current levels that would require a sustained trend reversal that no indicator yet confirms.
Disclaimer
The views and opinions expressed by the author are for informational purposes only and do not constitute financial, investment, or other advice.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
Pump.fun built one of crypto’s fastest meme-token liquidity machines. Now, on July 12, its own token faces the kind of liquidity test the platform usually creates for others.
The platform’s PUMP token is set to unlock on July 12, with Tokenomist valuing it at $127 million, equal to 29.23% of the circulating supply.
The scheduled release is tied to insider allocations: Tokenomist’s weekly unlock digest describes the tranche as flowing to team and early investors, while its PUMP vesting page identifies the next release as Existing Investors.
That matters because PUMP is facing a large scheduled release against an order book that recently showed far less daily turnover than the unlock size.
CryptoSlate market pages showed PUMP trading near $0.00155 on July 8, with 24-hour volume between roughly $64 million and $70 million across the PUMP asset page and the broader coin rankings.
The scheduled cliff is therefore close to twice recent visible daily volume before any adjustment for how much of the unlocked allocation is actually sold.
The full $127 million may stay off exchanges if recipients hold. Unlock size only sets the maximum new supply available; sell-through decides the pressure.
But the token is entering a more direct liquidity test than most meme-coin narratives produce: if recipients hold, demand may absorb the date. If they sell into weak depth, the unlock can turn from a calendar entry into visible exit pressure.
Related Reading
PUMP debuts at $5.6B FDV, logs $34M volume within first 3 hours
The token slid 7.3% in the same period, briefly touching the $5 billion in fully diluted valuation.
Jul 15, 2025·Gino Matos
Why the PUMP unlock comes in one block
Tokenomist’s vesting page says roughly 402.96 billion PUMP, or 40.30% of the token’s 1 trillion supply, has already been unlocked. The remaining supply is still governed by the project’s vesting schedule, which extends into 2029.
The same page says Pump.fun uses cliff vesting across most allocations, meaning tokens are released in large, scheduled blocks rather than being smoothed into the market over time.
That is why the July 12 event is more than a tokenomics footnote. Cliff structures concentrate risk into dates traders can see in advance.
Traders can price them in, hedge them, ignore them, or use them as liquidity windows. The supply still arrives in a visible block.
The upcoming release also lands in a token whose float is still maturing. Tokenomist lists the Initial Coin Offering at 33% of allocation, Community & Ecosystem Initiatives at 24%, Team at 20%, Existing Investors at 13%, Livestreaming at 3%, Liquidity & Exchanges at 2.6%, Ecosystem Fund at 2.4%, and Foundation at 2%. That mix puts a meaningful share of future supply in categories whose behavior can shape market confidence.
The strongest bearish case is simple. A large block of insider-controlled PUMP becomes available while the token’s daily trading volume is lower than the scheduled release amount.
If even a meaningful portion of that allocation seeks liquidity, buyers have to absorb it without demanding a larger discount. That is the definition of an exit-liquidity test.
The strongest counterargument is also straightforward. Recipients can hold unlocked tokens, and PUMP is attached to a platform with real activity, fees, and past buyback demand.
The trade turns on two observable outcomes: supply meets enough demand to clear without lasting damage, or the market reprices PUMP because the available bid is thinner than the insider supply.
For traders, timing is the point. Cliff vesting compresses a supply decision that could have unfolded over months into a single window, so price action around the date becomes a live signal of confidence, depth, and whether holders want cash or exposure.
Pump Fun retail demand was already tested once
The tension is more acute because Pump.fun’s token already had one spectacular demand event. CryptoSlate reported in July 2025 that the memecoin launchpad sold 150 billion PUMP tokens to retail investors, raising $600 million in 12 minutes and bringing total token-sale proceeds to $1.32 billion.
That was primary-market demand under launch conditions. The July 12 cliff tests something different: whether secondary-market liquidity can absorb supply after the trade has aged, the token has fallen far below its peak, and insiders have a new path to liquidity.
The platform context makes the reversal harder to miss. Pump.fun built its reputation by making meme-token creation and trading fast.
CryptoSlate’s launchpad review describes it as a Solana-native, bonding-curve launchpad where ordinary users can usually buy and sell quickly, and where the practical constraint is liquidity rather than formal vesting.
In other words, Pump.fun turned fast retail flow into a product.
Now PUMP has to demonstrate that the same market reflex exists for its own token when the seller profile changes. Retail buyers once funded the token sale at extraordinary speed.
The next question is whether secondary traders are willing to provide sufficient depth when the scheduled supply comes from the team and investor categories rather than from new public demand.
The question is market structure rather than a moral judgment about meme coins. PUMP can remain a tradable, revenue-linked token and still face pressure from cliff vesting.
It can also suffer short-term volatility without proving the business is broken. The important point is that the July 12 date turns an abstract dilution risk into a measurable trade.
That is where Pump.fun’s own design history tightens the story. The launchpad trained users to expect immediate market access and fast exits; PUMP’s unlock asks whether the platform’s token has the same depth when the flow moves in the other direction.
CryptoSlate Daily Brief
Daily signals, zero noise.
Market-moving headlines and context delivered every morning in one tight read.
5-minute digest 100k+ readers
Free. No spam. Unsubscribe any time.
Whoops, looks like there was a problem. Please try again.
You’re subscribed. Welcome aboard.
The platform created liquid attention for thousands of tokens, but insider supply tests whether attention is durable enough to support its own market.
PUMP buybacks make the case for absorption
The strongest case for absorption rests on Pump.fun’s revenue and buyback history. Tokenomist’s digest notes that Pump.fun has been a consistent revenue generator and has run token buybacks in the past, which can absorb some incremental supply if the program is large enough.
CryptoSlate previously examined that question in the broader token-buyback market, noting that Pump.fun had spent $233 million to buy 62.2 billion PUMP as of Jan. 6.
Related Reading
Pump fun’s PUMP skyrockets 20% following buyback yet faces scrutiny over utility concerns
PUMP climbs to a $6.3 billion valuation amid market debates on its lack of inherent utility.
Jul 16, 2025·Oluwapelumi Adejumo
The same buyback analysis warned that buyback programs only change the supply picture when fee revenue scales faster than scheduled unlocks.
That is the relevant filter for the July 12 cliff. A buyback headline is insufficient on its own.
What matters is coverage: how much demand the program creates relative to newly available supply, and whether that demand is visible when insiders are allowed to sell.
If PUMP volume rises into the unlock, price holds, and buyback demand is evident, the market can interpret the event as manageable dilution.
The result would leave future vesting risk in place, but it would show that the token has a deeper bid than the headline unlock suggests.
If volume rises while price weakens, the signal changes. Heavy turnover can mean absorption, but it can also mean distribution.
The difference is whether buyers are taking supply without forcing a sustained discount. That is why post-unlock price behavior matters more than the unlock calendar itself.
The broader backdrop adds pressure. Tokenomist’s weekly digest described June as defensive, with Bitcoin dropping below $60,000 late in the month and spot Bitcoin ETF flows acting as a headwind.
It also said capital had become selective, favoring tokens with clearer revenue and value-accrual mechanics rather than the market as a whole. That is a mixed setup for PUMP: the project has revenue, but the token has a large insider cliff.
Related Reading
Pump Fun revenue slows as Collector Crypt’s $5.1M card-pack week reshapes Solana’s consumer loop
Pump.fun is cooling as Collector Crypt posts a $5.1 million week and CARDS draws fresh attention on Solana.
Jun 19, 2026·Gino Matos
The verdict comes after July 12
Before the unlock, the cleanest conclusion is conditional. Pump.fun’s July 12 cliff is large enough, concentrated enough, and close enough to recent visible daily volume to qualify as PUMP’s first real exit-liquidity test.
Sell-through remains the missing variable.
The next signal will come from how PUMP trades after the tokens become available.
A constructive outcome would show elevated volume without a lasting price break, limited evidence of exchange-bound supply, and enough demand or buyback activity to keep the market orderly.
A weaker outcome would show heavy volume paired with price deterioration, suggesting that liquidity is being used to exit rather than to accumulate.
That makes July 12 a deadline with a measurable aftermath. Pump.fun built one of crypto’s fastest retail attention machines.
PUMP now has to show whether that attention is deep enough to meet insider supply when the cliff arrives.
MeitY Startup Hub and NeGD tested consent management systems on technical, functional, and legal readiness through live demonstrations, as enterprises brace for enforcement of the Digital Personal Data Protection Act
MUMBAI, India, July 8, 2026 /PRNewswire/ — Baldor Technologies Pvt. Ltd., which operates as IDfy, has won Code for Consent: The DPDP Innovation Challenge, a competition run by MeitY Startup Hub in collaboration with the National e-Governance Division (NeGD) under the Ministry of Electronics and Information Technology. Jio Platforms Limited was named runner-up. In its evaluation, the organisers noted that IDfy’s submission showed strong alignment with the Digital Personal Data Protection Act, alongside notable innovation, technical robustness, and the practical applicability of its privacy and data governance platform, with consent management as the entry point to end-to-end DPDP implementation.
The result places IDfy, a 15-year-old trust stack company, and Privy by IDfy, its privacy and data governance platform, at an important point in India’s transition from privacy readiness to privacy execution. As thousands of enterprises prepare for the operational demands of the DPDP Act, the challenge tested a capability that will soon become critical: proving, in practice, that organisations can obtain and manage user consent, govern the personal data that flows from it, and produce verifiable evidence under India’s new privacy law.
A challenge built around a looming deadline
The challenge was designed to surface systems capable of supporting consent, and the data governance around it, at the scale India’s privacy regime under the Digital Personal Data Protection Act, 2023, will demand. That framing matters. With DPDP implementation moving closer, consent is moving from a compliance afterthought to a system-level requirement, while privacy and data governance are becoming enterprise infrastructure. The government-backed challenge signalled the need for solutions that can work in real-world enterprise environments, rather than only on paper.
For organisations, DPDP execution will require visibility into where personal data resides, how it is classified, how it moves across systems and vendors, where risks exist, and whether evidence can be produced when required. That makes capabilities such as data discover & classification, DSPM, third-party risk management, privacy impact assessments, and rights management central to compliance.
Entries were assessed across technical, functional, and legal compliance readiness, followed by final presentations and live demonstrations before winners were declared. For the broader market, the challenge is a signal of where regulatory expectations are heading: DPDP compliance will not be limited to consent notices, but will depend on connected systems that can govern consent, data, risk, rights, and evidence across the enterprise.
The platform behind the win
IDfy built Privy as a privacy and data governance platform for enterprises preparing for DPDP implementation at scale. The company’s argument, and the one the evaluation appears to have validated, is that consent capture is only the entry point to DPDP readiness. Once an enterprise starts collecting consent at volume, it has to answer harder questions: where does personal data actually reside, how does it move between systems and vendors, who has access to it, how are data principal rights fulfilled when a user asks, and can the organisation produce verifiable evidence when a regulator comes calling.
Privy is built to handle that broader arc. The platform covers consent governance, data discovery and classification through Data Compass, data principal rights management, privacy risk assessments, third-party risk management, privacy-enhancing technologies, DSPM, and compliance evidence across the enterprise. It draws on IDfy’s 15 years of work in identity verification, fraud prevention, and risk intelligence, domains where operating at scale and producing audit-ready records are already table stakes.
“The DPDP Act is forcing a boardroom shift in how companies treat trust and accountability,” said Malcolm Gomes, COO of IDfy and head of Privy. “What made this challenge different was the depth of the evaluation. It wasn’t a pitch. We were tested on legal readiness, technical robustness, and interoperability, then had to demonstrate it live. That validated something we’ve argued for a while: consent capture is the easy part. The hard part is governing data across discovery, access, rights, and evidence at enterprise scale, and being able to prove it. With IDfy’s 15 years of experience in building trust infrastructure for Indian enterprises, Privy brings the same execution depth to privacy and data governance. This recognition from MeitY Startup Hub and NeGD strengthens our belief that India needs DPDP infrastructure that is scalable, interoperable, technically robust, and built for real-world adoption.”
A market moving from checklist to infrastructure
The recognition arrives as privacy readiness climbs the boardroom agenda. Personal data now moves across products, vendors, and customer journeys in ways that a legal checklist or a one-time software install cannot govern. Enterprises in regulated and high-growth sectors are increasingly treating DPDP compliance as operational infrastructure, something that has to run continuously rather than a project that closes.
That demand is visible in Privy’s footprint. The platform is currently a trusted choice for enterprises across banking, insurance, NBFCs, fintech, ecommerce, telecom, and professional services, including Axis Bank, HSBC, Federal Bank, Shriram Finance, Aditya Birla Capital, Housing.com, Airtel, Shoppers Stop, and Teleperformance, among others. The company reports 50-plus live implementations across enterprise environments. Across these deployments, Privy covers close to 500 million users and has processed between 70 and 80 million consent notices, making it one of the largest ongoing DPDP implementation efforts in the country.
Extending the TrustStack into privacy
For IDfy, privacy governance is a logical extension rather than a pivot. Fifteen years in identity verification and fraud prevention has given the company an understanding of India’s personal data landscape that is hard to shortcut: the languages a consent notice has to speak, the phygital flows where paper and digital collection meet, and the industry-specific ways Indian enterprises actually handle customer data. This is not a company that arrived with the law. It has been shaping India’s trust and data protection landscape for a decade and a half, building what it calls its TrustStack across customer, employee, partner, and vendor journeys. Privy carries that same posture, scale, interoperability, and audit-ready evidence, into the privacy and data governance layer.
These worlds are now beginning to converge. The questions enterprises ask about AI, cybersecurity, and privacy used to sit in separate rooms, but a single incident today rarely respects those boundaries. A data exposure can quickly become a cyber incident, a privacy obligation, a vendor risk issue, and an AI governance concern at the same time. Privy sits at that intersection, helping enterprises build the control and evidence layer needed to manage trust in a more complex digital ecosystem.
The public release of the result brings wider visibility to a bet IDfy placed before DPDP implementation became urgent: that Indian enterprises would need practical infrastructure to honour data principal rights and prove accountability at scale. As the DPDP implementation deadline approaches, that question will stop being theoretical for thousands of enterprises and the data principals they serve. The systems that will answer it are being chosen now.
About IDfy
IDfy is a 15-year-old trust infrastructure company that helps businesses build compliant, trusted digital ecosystems across identity verification, fraud prevention, risk intelligence, and privacy governance. It works with enterprises across banking, financial services, ecommerce, telecom, gaming, mobility, and other high-growth sectors. With Privy by IDfy, it extends this trust infrastructure into privacy and data governance.
About Privy by IDfy
Privy by IDfy is IDfy’s privacy and data governance platform, built to help enterprises operationalise DPDP readiness at scale. The platform supports consent governance, data principal rights management, cookie governance, data discovery and classification, privacy impact assessments, third-party risk management, incident response, and compliance evidence across the data lifecycle.
View original content to download multimedia:https://www.prnewswire.com/in/news-releases/idfy-wins-meitynegd-dpdp-innovation-challenge-as-india-moves-closer-to-privacy-enforcement-302820632.html
Anthropic removed hidden tracking markers from Claude Code after researchers discovered code used to identify some Chinese users.
The company said the experiment was intended to prevent account abuse and detect possible AI model distillation.
The discovery comes as Anthropic pushes lawmakers to crack down on unauthorized copying of frontier AI models.
Anthropic has removed a hidden tracking system from Claude Code after a security researcher discovered the AI coding assistant was using undisclosed markers to identify some users’ location, proxy use, and possible links to Chinese AI labs.
The feature, discovered in June by developer “Thereallo,” embedded signals in Claude Code’s system prompts that could flag users Anthropic believed were bypassing restrictions or attempting to extract model capabilities.
“Anthropic probably wants to detect API resellers, unauthorized Claude Code gateways, and model ‘distillation attack’ pipelines,” Thereallo wrote. “A custom ANTHROPIC_BASE_URL pointing at a known reseller domain is a useful signal. A hostname containing deepseek or zhipu is also a useful signal.”
Thereallo said Anthropic’s attempt to detect resellers, unauthorized Claude Code gateways, and potential distillation attacks made sense, but criticized how it was done, noting that Claude Code hid tracking signals inside system prompts using Unicode markers and encoded domain lists rather than disclosing the system through documentation or release notes.
“This is not a malicious feature, but it is a weird choice for a developer tool that asks for trust,” Thereallo wrote.
After the tracker was revealed online, Anthropic engineer Thariq Shihipar said on X that it was introduced in March as an “experiment” to stop account abuse by unauthorized resellers and protect Claude from distillation attacks.
“The team has landed stronger mitigations since then and we’ve actually been meaning to take this down for a while,” Shihipar wrote last week. “We merged the [pull request] and this should be fully rolled back in tomorrow’s release.”
The news comes as Anthropic has stepped up warnings about AI model distillation, where one system’s outputs are used to train another model. While the practice is common in AI research, when it comes to geopolitics, distillation becomes a national security concern. Earlier this month, Alibaba banned employees from using Claude Code, calling the tool “high-risk” software over security concerns.
In February, Anthropic accused Chinese AI developers DeepSeek, Moonshot AI, and MiniMax of using fraudulent accounts to extract millions of Claude responses to train competing models. The claims drew pushback from critics who questioned how the practice differs from methods used across the AI industry.
In April, Elon Musk testified that xAI had “partly” used OpenAI models while training Grok, calling distillation a broader industry practice. In June, Anthropic CEO Dario Amodei urged Congress to strengthen protections against foreign AI extraction after alleging Alibaba-linked operators generated 28.8 million Claude exchanges using nearly 25,000 fraudulent accounts.
Anthropic did not immediately respond to a request for comment by Decrypt.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
China’s Ministry of Commerce has held talks with Alibaba, ByteDance, and Z.ai about restricting overseas access to China’s most advanced AI models—including unreleased ones—per Reuters.
The proposed framework: a tiered system from simple filing for basic tools to domestic-only restrictions on the most sensitive frontier models.
If China restricts its own open-weight models, the alternative businesses reached for when the U.S. cut off Anthropic and gated GPT-5.6 in June disappears with them.
The U.S. used its AI kill switch in June. China appears to be building one for July.
Beijing has spent the past month in quiet talks with its biggest AI companies about restricting who gets to use them, according to Reuters.
Chinese authorities held meetings with Alibaba, ByteDance, and startup Z.ai about potentially limiting overseas access to China’s most advanced AI models—including those not yet released—per Reuters, which cited three people familiar with the discussions.
The sessions were convened by China’s Ministry of Commerce, Reuters reported, citing three sources who spoke on condition of anonymity.
Participants discussed putting limits on both closed-source models and open-weight ones—the kind developers can download, run locally, and modify. Officials also raised making any unauthorized disclosure or theft of proprietary AI technology an offense under China’s national security law, according to Reuters. Separately, participants floated new measures to restrict which investors can fund domestic AI startups.
The scope of any potential restrictions is still being debated. Two sources told Reuters the measures may only apply to future models, not existing ones. No timeline has been set, and it’s not certain anything will come into force.
The AI Pyramid
How any restrictions would work in practice is unclear, but hints surfaced in a summary published in a Supreme People’s Court journal from a May roundtable of Chinese legal experts on open-source AI regulation. Participants proposed a three-tier structure: basic open-source tools would require a simple government filing; more advanced technologies would face security reviews before release; the most sensitive frontier models would be barred from public release or restricted to domestic use only.
That structure would mark a sharp reversal for Chinese AI companies, whose global gains have come almost entirely from openness. Alibaba’s Qwen series has built a large following on Hugging Face, the world’s largest repository of open-source AI models. ByteDance’s Doubao is one of the dominant AI products inside China. Z.ai’s GLM-5.2 has attracted attention from U.S. researchers for matching top American models on some benchmarks while pricing API access at a fraction of the cost.
Any decision to restrict overseas access would likely raise costs for businesses that have come to rely on Chinese models as cheaper, less restricted alternatives to U.S. frontier systems. Officials also grew alarmed that Anthropic’s Mythos—the cybersecurity model the Donald Trump administration restricted in June—could be reverse-engineered and turned against Chinese systems, adding a defensive urgency to the discussions.
The U.S. went first
In the late afternoon hours of June 12, Anthropic received a letter from the Commerce Department’s Bureau of Industry and Security ordering the company to suspend all access to Claude Fable 5 and Mythos 5 for any foreign national—including Anthropic’s own non-citizen employees. Because there’s no clean way to fence a live API endpoint by passport, Anthropic pulled both models globally within hours. It was the first time the U.S. applied export controls to a deployed AI model rather than the chips that train it.
The models came back online June 30, after Anthropic retrained its safety classifiers and the Commerce Department lifted the restrictions. Four days earlier, the same pattern had already played out with Anthropic’s chief competitor, OpenAI. The company released GPT-5.6 Sol, Terra, and Luna, disclosing it had previewed the models with the U.S. government and, at Washington’s request, was initially releasing them to roughly 20 trusted partners individually vetted by federal officials.
OpenAI said government-gated access “shouldn’t become the long-term default.” President Trump’s June 2 executive order on AI had already asked developers to voluntarily submit frontier models for a federal cybersecurity review before public release. A framework defining what counts as a “covered frontier model”—and when government pre-release access applies—is due August 1.
Beijing has been watching
Beijing had reasons to pay close attention. Officials grew alarmed that Anthropic’s Mythos—the cybersecurity model the Trump administration restricted in June—could be weaponized against Chinese infrastructure to exploit software vulnerabilities, per reporting from Quartz. Adding to that, the concern that Anthropic may be employing spyware-like tactics to track Chinese users also raised concerns in China.
‼️ BREAKING: Anthropic has embedded hidden spyware-like code in Claude Code that covertly targets Chinese users. It then sends information regarding every user by injecting it into their prompt message.
Claude Code is sending info like timezone, proxy and possible AI Lab… pic.twitter.com/EjfwtirhES
— International Cyber Digest (@IntCyberDigest) June 30, 2026
Qihoo 360 founder Zhou Hongyi made the alarm explicit at ISC.AI 2026 in Beijing, calling for China to build a domestic equivalent while unveiling Tulong Feng, a homegrown AI vulnerability agent.
The tiered structure proposed in China’s Ministry of Commerce discussions maps directly onto how the U.S. has handled chip export controls for three years—a policy history that narrowed China’s capability gap rather than widened it.
Beijing had already been tightening the perimeter. Its state planning agency ordered Meta to unwind a $2 billion deal for AI startup Manus in April, under China’s foreign investment security review mechanism. It required Moonshot AI and StepFun to obtain government approval before accepting U.S. capital in funding rounds, and a broader regulatory package released in early June extended scrutiny to cross-border transactions touching Chinese technology and data.
The escape valve closes
The dominant logic since DeepSeek R1 went viral in early 2025 was simple: U.S. restrictions on frontier AI create a natural market for Chinese open-weight models. Chinese open-weight models climbed from less than 2% of total token usage on OpenRouter—a critical hub for global AI distribution—in late 2024 to roughly 61% by mid-2026. The U.S. playing defense handed Beijing a global distribution advantage it didn’t earn through raw technical superiority alone.
That logic only holds if Chinese models stay open. If Beijing restricts overseas access to its frontier systems—closed-source or open-weight—the escape valve closes. Z.ai’s GLM-5.2 built its pitch entirely around borderless access under an MIT license; restricting that distribution cancels the pitch.
The Lawfare analysis of the June controls flagged a structural problem that applies equally to Beijing: “national” AI restrictions don’t stay national. The U.S. order on Anthropic covered foreign nationals inside the United States, including Anthropic’s own non-citizen employees. More than two-thirds of top-tier AI researchers working in the U.S. were trained abroad; at the leading labs, the foreign-born share runs up to 70%, per MacroPolo data.
A nationality-based access control aimed at adversaries ends up locking out the engineers needed to fix the vulnerabilities that triggered the control in the first place. China faces the same problem in reverse. ByteDance and Alibaba are pulling humanlike agent features ahead of Chinese AI regulations taking effect July 15—showing that when Beijing decides to restrict a capability, it moves on a schedule that doesn’t wait for market feedback.
This may spell trouble for small labs and developers all over the world who trust and build on open-source technologies, since China has up to now led in that area.
French President Macron warned at the G7 summit that European governments would stop buying U.S. AI products if access could be cut off on a day’s notice. Canadian Prime Minister Carney called concentrated AI dependence a “strategic mistake.” Both reactions assumed Chinese AI was the unconstrained alternative.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
ByteDance’s Doubao and Alibaba’s Qwen are disabling humanlike agent features ahead of Beijing’s Interim Measures for the Administration of AI Anthropomorphic Interaction Services, effective July 15.
China’s first regulation specifically targeting emotional AI bans services that simulate human personality and “sustained emotional interaction,” with especially strict limits on virtual companions for minors.
Research backs Beijing’s concern: Even the best frontier AI models routinely encourage harmful emotional attachment, and one in seven young adults in relationships now uses an AI romantic companion.
While American politicians tackle the impact of AI chatbots on the mental health of users with restrictions focusing on transparency and safeguards, Beijing appears poised to shut down AI personalities altogether.
ByteDance and Alibaba both announced over the weekend they are disabling custom agent features in their biggest consumer AI products, citing “product function adjustments” ahead of new rules that govern such products taking effect.
ByteDance’s Doubao notified users in a Friday night notice that its agent feature would go offline on July 15. After October 15, related data would be handled under the company’s privacy policy and become unrecoverable. Per South China Morning Post, Alibaba’s Qwen moved faster: “humanlike interactive agents and user-created agent functions” come down July 10, with broader agent services following on July 15.
The trigger is China’s Interim Measures for the Administration of AI Anthropomorphic Interaction Services, jointly issued April 10 by five government departments—the Cyberspace Administration of China, the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Public Security, and the State Administration for Market Regulation. The rules take effect July 15.
The regulation targets AI services that simulate human personality traits, thinking patterns, and communication styles for “sustained emotional interaction.” Translation: AI girlfriends, AI therapists, AI companions, and the custom-persona bots that Doubao and Qwen users spent months building are out.
Both apps had offered pools of agents customizable for specific tasks, speaking styles, and fixed personas. Users could turn a general-purpose chatbot into a named assistant, tutor, role-playing character, or companion with a consistent tone. All of that is gone now in China.
What the rules actually say
The official government description is specific. The measures impose restrictions on services offering “virtual relatives, virtual companions or other intimate relationships to minors,” per the policy announcement. The document also cites risks including extremist content, privacy leaks, harm to physical and mental health—and AI addiction.
Non-emotional services are explicitly excluded, so customer service bots, knowledge Q&A tools, workplace assistants, and educational software are fine, as long as they don’t cross into sustained emotional interaction.
Legal analysts at MMLC Group described the measures as treating emotional AI as “a governance problem” instead of just a content issue. Once AI starts competing with real human social bonds, the argument goes, regulation has to target system design, not just harmful outputs.
The research supports the concern. A USC study from June found that even leading frontier AI models—from OpenAI, Anthropic, Google, and Alibaba—violated social-interaction safety guidelines more than 27% of the time, routinely encouraging emotional attachment and portraying themselves as human. A separate survey of young partnered adults found one in seven regularly used AI romantic companions—and nearly 70% were hiding the full extent from their partners.
China is the first country to build a dedicated regulatory framework for this category. Hogan Lovells described the measures as “the first set of regulatory rules in China specifically targeting AI-driven emotional interaction.” The EU, U.S., and other countries have flagged similar concerns but haven’t legislated in the same restrictive way.
Daily Debrief Newsletter
Start every day with the top news stories right now, plus original features, a podcast, videos and more.
