Crypto World
Ethena surges as buyback vote, VC unlock overhaul boost token outlook

The changes aim to cut investor selling pressure and channel protocol revenue to ENA as Ethena looks to revive USDe growth.
Crypto World
Genius Group plans $827M Bitcoin, $800M AI treasuries
Genius Group has proposed raising capital through perpetual preferred securities to build an $827 million Bitcoin treasury and an $800 million AI portfolio within a $2 billion total-asset target for fiscal 2031.
Summary
- An initial preferred securities offering would seek $12.5 million from income-focused investors.
- Proceeds would be divided among Bitcoin, AI investments, and an 18-month dividend reserve.
- Genius Group plans to restart Bitcoin purchases after liquidating its holdings to repay $8.5 million.
- Final offering terms remain subject to board approval, securities rules, and market conditions.
Genius Group turns to preferred capital
Genius Group said on Aug. 27 that it intends to use its $1.2 billion shelf registration to issue publicly registered perpetual preferred securities, placing the financing method at the center of its five-year treasury plan.
The Securities and Exchange Commission declared the shelf registration effective on July 18, 2025. Such a registration allows the NYSE American-listed company to offer securities over time, but it does not represent SEC approval of the securities or their investment merits.
Under the preliminary proposal, Genius Group would seek $12.5 million in its first preferred securities offering. The company expects the instruments to be non-convertible and to carry a variable dividend paid monthly.
Funds from the sale would be allocated among the Bitcoin treasury, the AI treasury, and a U.S. dollar reserve equal to approximately 18 months of preferred dividend payments. Genius Group did not disclose how much of the initial proceeds each party would receive.
Discussions have begun with investment banks that have experience in preferred securities and digital asset treasury financing. However, the final issue price, dividend rate, offering size, exchange listing, and sale date have not been decided.
Any offering would require separate materials filed with or furnished to the SEC. The structure will also depend on board approval, applicable securities laws, regulatory requirements, and market conditions, according to the announcement.
Shareholders provided some of the required corporate authority at Genius Group’s annual meeting in July. About 97.58% of votes supported giving the board authority to issue preferred shares, while 99.54% approved a mandate allowing the company to repurchase up to 20% of its ordinary shares.
The $2B target covers total assets
Rather than placing $2 billion entirely into Bitcoin and AI investments, Genius Group has set separate fiscal 2031 targets of $827 million for its Bitcoin treasury and $800 million for its AI portfolio.
Operating businesses, cash, and other holdings would account for the remaining assets under the $2 billion plan. The company currently reports net assets of $106.6 million, following a 57% year-over-year increase announced on Aug. 13.
Genius Group calculated its net asset value at $0.62 per ordinary share. With GNS closing at $0.18 on Aug. 26, the company said its stock was trading at approximately 0.29 times book value, compared with what it described as a 2.60-times average for the U.S. education sector.
Management has forecast that net asset value could reach between $2 and $4 per share over five years if the company executes its financing, asset-purchase, and share-buyback plans. The projection also depends on market conditions and the performance of Bitcoin and its AI investments.
Chief executive Roger James Hamilton described perpetual preferred capital as a way to fund treasury purchases without issuing more ordinary shares.
“Every dollar of preferred capital deployed into our Bitcoin and AI Treasury that generates returns above the preferred dividend rate flows directly to our ordinary shareholders’ net asset value.”
Returns above the dividend cost could add to the assets attributable to ordinary investors. If the acquired assets lose value or earn less than the dividend rate, however, the preferred payment obligations would remain senior to ordinary shareholder distributions.
Genius Group identified Bitcoin price volatility, changes in private technology company valuations, financing costs, and capital availability among the factors that could cause actual results to differ from its forecasts.
Strategy provides the financing model
For its proposed securities, Genius Group has taken Strategy’s Bitcoin financing program as its main reference. The company said Strategy has raised more than $16 billion through four perpetual preferred stock series since introducing STRK in January 2025.
The preferred securities have no fixed maturity and do not require repayment on a set date. Their dividends and senior claims still create costs that treasury assets must cover before any excess return reaches ordinary shareholders.
Investor demand has emerged for some of the products. In May, Strategy’s STRC security recorded $1.53 billion in daily trading volume, crypto.news reported, drawing attention to the use of dividend-paying stock to fund corporate Bitcoin holdings.
Market prices can also depart from the issue or liquidation value. STRC fell to an intraday low of $82.50 on June 18 before closing near $88.59, well below the approximately $100 level around which the security was designed to trade.
Strategy later used Bitcoin sales to support the preferred program. An Aug. 10 SEC filing showed that the company sold 1,690 BTC for $108.6 million between Aug. 3 and Aug. 9, using the proceeds to repurchase about 1.15 million STRC shares.
A subsequent filing showed Strategy spent $132.2 million on additional STRC repurchases and $52.4 million on related dividends during the following week. It also placed $149.1 million into its U.S. dollar reserve, bringing the cash pool to $4.8 billion.
For U.S. investors, Genius Group’s final prospectus will determine the economic and legal terms of the proposed security. Until those documents are available, its dividend rate, liquidation preference, call provisions, exchange access, and possible tax treatment remain unconfirmed.
Bitcoin purchases are scheduled to restart
Before developing the dual-treasury plan, Genius Group pursued a Bitcoin-first policy under which it intended to hold at least 90% of its reserves in BTC.
The company adopted the policy in November 2024 and planned an initial $120 million purchase program. By January 2025, it held 420 BTC after buying another $5 million at an average price of $95,912 per coin. Holdings later reached a peak of 440 BTC.
A U.S. court order disrupted the program in early 2025 by restricting the company from selling shares, raising funds or buying Bitcoin during a legal dispute tied to its asset purchase agreement with Fatbrain AI. Genius Group reduced its Bitcoin holdings while seeking relief from the restrictions.
After the order was lifted, the company resumed purchases in June 2025 and increased its balance to 100 BTC. Management also restored a target of accumulating 1,000 BTC.
Liquidity needs later forced another change. Genius Group sold its remaining Bitcoin during the first quarter of 2026 and used the funds as part of the repayment of $8.5 million in debt.
Before the final sale, the company reported holding 84 BTC valued at approximately $5.7 million in March. Its April 1 operating update said it would rebuild the treasury when management considered market conditions more favorable.
Under the latest timetable, Bitcoin purchases are expected to restart in the fourth quarter of 2026. The company has not disclosed the size or price of its first planned acquisition.
Genius Group established the second part of its treasury in May 2026, when the board authorized an AI portfolio with an initial investment plan of up to $100 million. The company made its first allocation in June through funds providing exposure to private companies, including OpenAI, Anthropic, Anduril, and Databricks.
SpaceX held the largest look-through weighting at 13.5% of the AI portfolio, according to the company. Genius Group said its portfolio also contained exposure to xAI, Figure AI, Replit, and other companies involved in AI models, robotics, and infrastructure.
Crypto World
Grayscale Sees Zcash as Potential Bitcoin Challenger
Zcash could emerge as a meaningful challenger to Bitcoin’s dominance among digital assets as the rapid adoption of artificial intelligence puts a premium on financial privacy and fuels concerns over AI-powered surveillance, according to Grayscale.
In a new research report, Grayscale head of research Zach Pandl said Zcash (ZEC) has “second mover advantages” that could help it challenge Bitcoin’s (BTC) entrenched network effects, something previous alternatives such as Litecoin (LTC) have failed to achieve.
Central to Pandl’s argument is financial privacy. Zcash can shield transaction information, which Grayscale argues could become increasingly valuable as AI systems become better at analyzing financial activity at scale.
The report comes after ZEC’s roughly 19-fold increase over the past year. Despite those gains, Zcash remains valued at less than 1% of Bitcoin’s market capitalization, a disparity Grayscale sees as evidence of further upside if Zcash can capture market share.
Pandl acknowledged that Bitcoin’s liquidity and entrenched network remain powerful defenses of its dominant position. Grayscale also warned that Zcash remains a high-risk investment and that any further gains could be volatile and uneven.

Zcash could be valued at more than $4,000 if its market capitalization reached 5% of Bitcoin’s. Source: Grayscale
Related: Zcash’s Ironwood upgrade faces possible delay over infrastructure readiness
Zcash ecosystem attracts institutional capital
Interest in the Zcash ecosystem is broadening alongside ZEC’s strong price performance. As Cointelegraph recently reported, Nasdaq-listed privacy technology company Cypherpunk Technologies expanded its Zcash exposure by acquiring a mining fleet from Winklevoss Capital in a $33.33 million equity-based transaction.
The operation is already online across US facilities, producing about 4.2 GSol/s of Equihash hashrate, or roughly 18% of the Zcash network’s total computing power. Cypherpunk said the deal made its mining arm the network’s largest active fleet.
Related: Strategy’s $66B Bitcoin machine hinges on capital markets, not BTC price: Report
Crypto World
The 100 Most Influential People in AI 2026
Since the release of the 2000 Oscar-winning movie about her, the name Erin Brockovich has been synonymous with grassroots activism against powerful institutions. This year, Brockovich has turned her attention to a new target: AI data centers. When Brockovich asked people online how they felt about the issue, she received a “flood” of impassioned responses, she wrote in May.
So Brockovich launched a tool to map data centers across the country, allowing community members to submit their own information about nearby data centers. That map has since received more than 9,000 reports and has become a repository for information, including about data centers’ energy usage, physical size, and economic impact. The site also tracks local legislative and community efforts to push back. (Similar efforts include the Data Center Proposal Tracker and Data Center Watch.)
While Brockovich has elevated all kinds of concerns, she herself is particularly concerned about data centers’ lack of transparency and public participation. “Transparency means notifying residents before decisions are made, not after,” she wrote on her blog in May. “It means elected officials who answer to their constituents first, not to the corporations seeking tax breaks and zoning variances.”
Crypto World
Nvidia shares surge 8% on earnings beat, lifting technology stocks and bitcoin

Nvidia’s earnings beat and strong outlook lifted technology stocks, bitcoin and AI infrastructure companies.
Crypto World
Anthropic IPO Could Come in September, But It Has a Massive Risk Factor
Anthropic plans to let early investors and staff sell stock in its upcoming market debut, according to a recent report by The Information. SpaceX gave its own backers no such option in June.
The prospectus should land soon after Labor Day on September 7. That document sets out the risks and the finances before anyone can buy.
Anthropic IPO Departs From Musk’s SpaceX Structure
Big listings can sell two kinds of stock:
- New shares raise money for the company.
- Existing shares pay early backers instead.
SpaceX sold only the first kind. Its pricing release covered 555,555,555 new shares at $135 each. Not one came from an existing holder.
Underwriters then took another 83,333,333. That brought the total to 638.9 million shares and roughly $86 billion, still the largest listing ever.
Every dollar went to the company, with the filing estimating that Musk kept about 82.4% of the voting power once trading began.
Anthropic has copied part of that design, as indicated in a recent report. As BeInCrypto reported, it described supervoting shares for founders, the same tool Musk used to hold control.
Letting insiders sell changes the other half:
- Backers get cash at the offer price.
- Buyers absorb more stock on day one.
Follow us on X to get the latest news as it happens
Longer Lockups Could Offset the Early Selling
A lockup bars insiders from selling for a set period after a debut. It shields a young stock from a flood of supply. SpaceX shows what the delay looks like, because on August 6, about 911.5 million insider shares became sellable at once. That topped the 638.9 million sold in June.
The tradable pool more than doubled overnight, rising from 4.9% of the company to 11.8%. The stock still closed up 6.1% that day.
Anthropic appears to want the smoother path. A sale inside the deal is priced and placed with buyers in advance. A lockup expiry is neither.
The company is weighing lockups longer than the norm. Insiders would take cash early, then wait longer for a second window.
The backers in line are also large, given Anthropic raised $65 billion in May at a $965 billion valuation, according to its own announcement. Altimeter, Dragoneer, Greenoaks and Sequoia led that round.
Sovereign money joined too, with Singapore’s state fund GIC co-leading alongside Capital Group and Coatue. Those are the names that would be selling.
They would sell at a far higher price, as Anthropic last reported revenue in May, when it said its run rate had exceeded $47 billion. It has not updated that number since.
The prospectus would answer some of these questions. It is also expected to name public backlash against AI as a formal risk.
Crypto traders already price the same stake. Anthropic exposure trades through pre-IPO token markets on Solana, where PreStocks handles 78% of OpenAI and Anthropic volume.
The filing will probably name who sells and for how much, with the list likely to reveal more about Anthropicis valuation.
The post Anthropic IPO Could Come in September, But It Has a Massive Risk Factor appeared first on BeInCrypto.
Crypto World
Tokenized deposits may lift US borrowing costs
Tokenized deposits could make bank funding less “sticky,” potentially increasing borrowing costs for US households and businesses, according to an analysis by economists at the Federal Reserve Bank of Dallas. The concern is not about immediate, one-for-one changes to lending, but about how faster deposit movement—enabled by instant settlement and automated transfers—could reshape how banks manage liquidity and credit risk.
In a research note, economists Rosie Levy and Srini Ramaswamy argue that programmable deposit tokens combined with automated transfer mechanisms could allow customers seeking higher yields to switch banks more quickly. They estimate that if deposits became 10% more responsive to interest rates, banks’ capacity to hold long-term loans and other assets could decline by roughly $700 billion on a 10-year-equivalent basis. A separate scenario where deposits stayed at banks 10% less time implies a reduction of about $580 billion, expressed in the same 10-year-equivalent terms. These are scenario outcomes, not forecasts.
Key takeaways
- Tokenized deposits may increase deposit “rate sensitivity,” making funding more mobile when higher yields appear elsewhere.
- Instant settlement and automated transfers could shorten how long deposits remain at a given bank, reducing stability.
- Dallas Fed researchers estimate large liquidity and balance-sheet capacity effects under two 10% sensitivity/time scenarios, though they are not direct lending cuts.
- Banks are already building shared blockchain-style networks intended to move tokenized deposits within the regulated banking system.
Why instant settlement could destabilize funding
Levy and Ramaswamy’s central mechanism is straightforward: when settlement happens instantly, customers can react to rate differences faster. In traditional banking, moving deposits can take time, which can blunt how quickly funds shift across institutions. With programmable deposit tokens, deposits can be designed to integrate with automated processes—potentially powered by agentic artificial intelligence—that coordinate transfers with less manual friction.
The economists describe this as a shift in deposit behavior: deposits become more sensitive to interest rates and potentially less time-bound at a single bank. That matters because bank lending relies on relatively stable funding to support longer-duration assets.
Importantly, the authors stress that their numerical estimates are scenario-based. The changes are framed in terms of banks’ capacity to hold long-term loans and other assets, not as a direct “dollar-for-dollar” reduction in lending.
What the Dallas Fed scenarios imply for banks and borrowers
Under one scenario, the researchers assume deposits become 10% more sensitive to interest rates. Under another, deposits remain at banks for 10% less time. In both cases, they estimate reductions in banks’ capacity to hold long-term assets—about $700 billion and $580 billion, respectively, using 10-year-equivalent measures.
The analysis points to trade-offs banks could face when deposit stability declines. One response could be holding larger portfolios of highly liquid assets, such as reserves and US Treasurys, to better withstand faster outflows. Another could be leaning more on term debt to maintain the lending book.
But both adjustments can come with costs. Increasing reliance on wholesale funding or term debt typically raises funding expenses, and those higher costs can propagate into credit terms for borrowers—precisely the outcome Levy and Ramaswamy say could increase credit costs for US households and businesses.
From research to rollout: bank networks for tokenized deposits
The Dallas Fed concerns arrive as US banks accelerate plans for tokenized-deposit infrastructure. On Tuesday, 39 US state banking associations formed the BankChain Alliance, aiming to develop a nationwide network designed to support tokenized deposits, stablecoins, and automated settlement. Separately, The Clearing House is developing another network backed by major institutions including JPMorgan Chase, Bank of America, Citi, BNY, and Wells Fargo.
Banks have also begun connecting tokenized-deposit systems across organizations. On Aug. 20, Standard Chartered and HSBC completed a live cross-border transaction through Swift’s blockchain ledger. The reported design linked the two banks’ separate systems and recorded obligations on the ledger, with settlement still occurring via existing payment infrastructure.
Taken together, these efforts suggest that the industry is moving beyond pilots toward interoperable systems. From a policy perspective, that raises a key question Levy and Ramaswamy implicitly put on the table: if the plumbing makes movement faster and more programmable, will regulators and banks anticipate and manage the resulting funding dynamics?
Liquidity lessons from instant payments—what’s comparable and what isn’t
Levy and Ramaswamy cite Brazil’s Pix instant-payment system as a comparison point, while emphasizing it is not identical to tokenized deposits. Their reasoning is that instant-payment rails can change how quickly funds can move, which can alter deposit behavior and, in turn, banks’ balance-sheet choices.
A 2025 study by Brazil’s central bank found that heavier Pix use increased banks’ holdings of liquid assets and reduced credit intermediation. While that evidence does not prove the same outcome will occur with tokenized deposits, it offers a relevant reference for how faster payment flows can influence bank liquidity decisions.
For investors and lenders, the policy takeaway is less about whether tokenization will “help or hurt” lending in the abstract and more about how institutions will adapt their asset-liability management. If deposit mobility rises, market participants should watch for shifts in liquidity buffers, reliance on wholesale funding, and credit pricing—channels the Dallas Fed analysis highlights.
Going forward, the key uncertainty is how quickly tokenized deposit networks translate into real consumer and business deposit switching behavior. Readers should watch for regulatory guidance around tokenized deposit frameworks and for measurable changes in banks’ funding structures—especially whether liquidity reserves and term-debt reliance rise as these systems expand.
Crypto World
Bitcoin’s 23% Surge Lifts Miners Above AI Stocks
Bitcoin’s August rebound is dragging attention back toward the companies most directly leveraged to the network, with beaten-down mining stocks sharply outperforming many firms that had emphasized artificial intelligence (AI) and high-performance computing (HPC) rather than pure crypto exposure.
According to BlocksBridge Consulting’s Miner Weekly, Bitcoin has rallied about 23% over the past week—an advance that the firm says has outpaced most AI-linked infrastructure stocks. While some AI/HPC peers managed modest gains, the strongest moves were seen among miners with the most direct relationship to Bitcoin price action.
Key takeaways
- BlocksBridge reports Bitcoin’s ~23% weekly rally outperformed many AI-linked infrastructure equities.
- Bitcoin-focused miners such as Canaan, American Bitcoin, and Cango rose between 41% and 67%, while several AI/HPC-heavy names were flat or down.
- BlocksBridge attributes the stock surge to Treasury liquidity-support buybacks, renewed US crypto regulatory momentum, and a sharp liquidation-driven short squeeze.
- The broader message for investors: despite the industry’s AI pivot, Bitcoin price still largely determines near-term mining-stock performance.
Miners rebound while AI/HPC pivots lag
The outperformance was concentrated in companies that market themselves around mining and Bitcoin economics, rather than broader compute infrastructure narratives. BlocksBridge highlighted three beaten-down Bitcoin mining companies—Canaan, American Bitcoin, and Cango—posting gains ranging from roughly 41% to 67% during the same period.
By comparison, miners and infrastructure-linked companies with more diversified exposure saw smaller moves. BlocksBridge cited CoreWeave rising about 21%, Nebius gaining about 17%, and IREN up around 15%. It also noted that some players with heavier AI/HPC exposure were flat or declined—an important contrast for traders who may have been expecting the market to continue rewarding the AI theme alone.
This relative performance matters because it signals how quickly capital can rotate back toward the most straightforward “beta to Bitcoin.” Even if AI-related businesses remain central to longer-term strategy for many operators, the market appeared to treat Bitcoin price strength as the dominant driver of miners’ equity re-rating over the week.
Why Bitcoin’s rally translated into stock gains
BlocksBridge pointed to three overlapping catalysts that helped ignite the move in crypto markets and, in turn, mining equities.
Treasury buyback expansion and liquidity expectations
One driver was the US Treasury Department’s Aug. 19 announcement that it would at least double the size of its liquidity-support buybacks for longer-dated Treasury securities. BlocksBridge linked this to improved liquidity expectations—an environment that often supports risk assets broadly, including crypto-linked equities.
The connection is straightforward: when liquidity conditions improve, markets tend to become more willing to price risk higher, which can benefit volatile sectors like crypto and crypto mining.
Regulatory optimism after White House meeting
A second catalyst cited by BlocksBridge was renewed regulatory optimism following a White House meeting with crypto executives. Reuters reported that US President Donald Trump urged Congress to pass a “fair version” of the CLARITY Act, a stalled crypto market structure bill.
For public miners, regulation matters less as a daily operational variable and more as a factor that can influence investor confidence and capital allocation. Even the anticipation of clearer market rules can change how equity markets discount regulatory risk across the crypto complex.
Liquidations and a short squeeze after Bitcoin broke out
The third element described by BlocksBridge was a sharp short squeeze after Bitcoin’s breakout. The firm said more than $1.6 billion in crypto positions were liquidated over 24 hours—an event that can force margin closures, accelerate price momentum, and pull additional participants into the trade.
This type of positioning-driven rally can especially benefit miners’ stocks in the short term. Mining equities often move in tandem with both crypto prices and broader risk appetite, so liquidation cascades can amplify gains beyond what “fundamentals” alone would suggest.
Bitcoin still sets the pace, despite the AI funding push
BlocksBridge’s findings align with earlier coverage from Cointelegraph that Bitcoin’s rally lifted crypto-related stocks, including Bitcoin miners. While the industry has increasingly marketed AI and HPC capabilities in recent years, the week’s performance suggested that investors were still willing to pay up for direct exposure to Bitcoin rather than compute-adjacent narratives.
That focus is reinforced by additional BlocksBridge analysis referenced in the report: publicly traded Bitcoin miners have invested roughly $15 into AI data centers for every $1 in AI-related revenue generated. The firm cited nine public miners generating $341.2 million in AI and HPC revenue so far in 2026, alongside $5.11 billion in AI and HPC capital expenditures on the technology.
The implication is not necessarily that AI pivots are failing, but that the investment phase may be heavy and slow to convert into revenue. In such a setup, equities can be more sensitive to near-term Bitcoin price movements—because the market may not yet fully “see” AI returns in earnings, cash flow, or guidance.
What investors should watch next
If liquidity expectations, regulatory headlines, and crypto positioning remain supportive, mining stocks may continue to trade as a high-beta reflection of Bitcoin. But investors will likely watch whether the outperformance persists after the initial squeeze fades—and whether miners’ AI/HPC spending begins to translate into measurable revenue gains that can support valuations independently of Bitcoin’s next move.
Crypto World
Sovereign Digital Bond Used as Collateral in Onchain Repo
Virtu Financial, M1X Global and Tradeweb completed an onchain repo transaction using a sovereign digital bond as collateral, with the full transaction settling on the Canton Network.
The transaction used USDM1, a US dollar-denominated sovereign bond issued onchain by the Republic of the Marshall Islands and backed 1:1 by short-term US Treasurys. The bond pays a coupon while being used as collateral and is structured under New York law as a fully collateralized sovereign obligation.
Both companies said it was the first repo to combine natively issued sovereign collateral with fully onchain atomic settlement. Executed between regulated counterparties on Tradeweb, the full repo and repurchase cycle was completed in under 10 minutes.
The transaction puts tokenized sovereign debt to use as collateral in an institutional financing transaction, rather than solely as an asset for issuance or trading, though it remains an early-stage example and it is not yet clear whether the model will see broader adoption across institutional repo markets.
USDM1 is available through electronic trading platform Tradeweb, with institutional custody provided by Anchorage Digital, BitGo and tZERO, according to the release.
Related: Digital Asset lands $355M as a16z doubles down on Wall Street blockchain rails
Canton Network sees flurry of institutional activity
Canton is a blockchain network designed for institutional finance, with privacy and permissioning features aimed at regulated transactions and tokenized assets.
Thursday’s repo follows a July transaction in which Tradeweb facilitated the real-time transfer of a tokenized US Treasury from Franklin Templeton to Virtu Financial on Canton, settling the transaction against USDCx.
Network activity accelerated in August. FalconX and Interstice launched a cross-chain swap engine connecting Canton with Ethereum, Solana and Robinhood Chain, while World Liberty Financial launched its USD1 stablecoin natively on Canton.
Digital Asset and the American Idea Foundation, founded by former US House Speaker Paul Ryan, also announced plans this month for a 2027 pilot that would use Canton to distribute state-administered benefits across three US states.
Magazine: SHRINCS BIP published: Quantum-secure Bitcoin comes with a catch
Crypto World
The 100 Most Influential People in AI 2026
Joseph Gordon-Levitt sees his contribution to the AI conversation as a storyteller, helping people understand what widespread adoption of the technology and the current economic structures around it could mean for them. Companies, he notes, are incentivized to maximize shareholder value. “And that’s not necessarily going to line up with the technology being good for people,” he tells TIME.
As an actor and filmmaker, he’s also concerned about how AI will affect storytelling itself. In December 2025, he founded the Creators Coalition on AI alongside others including Natasha Lyonne and Daniel Kwan. The group aims to become Hollywood’s voice at the table, calling for fair compensation, job protections and deepfake guardrails, among other things. The coalition’s signatories include A-listers like Ben Affleck, Cate Blanchett, Margot Robbie, and Octavia Spencer.
Gordon-Levitt, who co-founded online collaborative media platform HitRecord in the early aughts, isn’t against the development of the technology itself, he says. He sees the potential for AI to empower more people. “But that’s not the path it’s on right now,” he says. “Right now it’s on a path where we take many steps back, where the power is taken away from more and more people and put into the hands of the few. It doesn’t have to be that way.”
In March, he was appointed as the U.N.’s first-ever global advocate for human-centric digital governance, where he would work with the organization’s Internet Governance Forum on promoting an approach that it hopes will be “equitable, innovative, responsible and human-centred.”
“Private companies can’t be the only ones deciding how this technology is going to be developed and deployed,” Gordon-Levitt says. “The people need to get involved. We can’t just be customers, we have to be advocates.”
Ever the storyteller, Gordon-Levitt is looking to distill these themes in his upcoming untitled thriller. “Talking about numbers and issues and policy and technology only takes a conversation so far,” he says. “Making a feature film is one of the best ways to really communicate a human experience—what it feels like as a person to be in this new world that we might be headed for.”
Crypto World
Who is Responsible When an AI Agent Loses Your Money?
On May 4, a message hidden in Morse code helped trigger a six-figure crypto transfer. It passed through two connected AI systems. One was Elon Musk’s Grok, the chatbot built by Elon Musk’s xAI. The other was Bankrbot, a crypto agent that could make payments from a linked wallet.
The attacker first sent the wallet a digital membership token that unlocked Bankr’s payment tools. Grok then decoded the message, and Bankrbot treated the response as a payment order. It transferred an estimated $150,000 to $200,000.
A Morse-Code Message Became a Six-Figure Payment
Now, why is this concerning? Because the case highlights a six-figure exploit involving just two AI agents. One AI produced text. Another treated it as permission to spend.
If we look at the scale of AI agentic payments today, such scenarios could be a nightmare for the future of Agentic Finance.
Keyrock counted 176 million on-chain agent payments worth $73 million through April 2026. The median payment sat between $0.01 and $0.10, while 76% fell below $0.30. Small payments become a large control problem when software can make them continuously.
The pattern is moving into mainstream payment infrastructure. Mastercard launched Agent Pay for Machines in June for high-frequency, low-value payments, while Google and Visa are developing standards for agents to prove identity and authority.
BeInCrypto asked Rodrigo Coelho, CEO of Edge & Node; Nitin Gaur, Head of Institutions at Nethermind; and Francesco Andreoli, Director of Developer Relations at MetaMask, who carries the risk.
Coelho was direct.
“The company that deployed it. There is no version of this where responsibility lands on the model,” said Rodrigo Coelho, the CEO of AI and Web3 infrastructure developer Edge & Node.
California has already put that principle into law. AB 316, effective since January, prevents a defendant who developed, modified, or used AI from arguing that the system autonomously caused the alleged harm. Causation and foreseeability still matter.
The Receipt Is Not the Permission
An on-chain transaction proves money moved. It does not prove the agent had a valid mandate to move it.
“Most companies deploying agents today could not actually prove what their agent was authorized to do. They can show you the transaction. It happened on a chain and the record is public and permanent. What they cannot show you is the permission that sat behind it,” said Coelho.
Gaps may include who delegated authority, which policy applied, what information the agent read and whether the payment stayed within its limits. A wallet address answers none of those questions.
Nitin Gaur from Nethermind said the dispute turns on the mandate.
“What decides a dispute is authority evidence. Show the agent acted inside a valid, signed, time-bounded mandate and this resolves like any other authorized payment.”
Google’s AP2 uses cryptographically signed mandates to record user intent. Visa’s Trusted Agent Protocol lets approved agents present digital signatures proving identity and associated authorization.
Mastercard adds credentialing and programmatically enforced limits. The rails differ, but the design goal is shared: permission has to travel with the payment.
Put the Limits Where the Agent Cannot Reach
A mandate still fails if the agent can rewrite it, approve its own request or hold unrestricted signing power. Coelho draws the boundary at the private key.
“The agent should not hold the keys. It should be able to propose a payment, and a separate system decides whether that payment is permitted,” said Coelho.
Francesco Andreoli from MetaMask makes the same point about prompts:
“The controls that work are the ones the agent cannot reach, if your policy lives in the prompt, it isn’t a policy, it’s a suggestion to a system we’ve repeatedly watched get talked into things.”
In practice, that means segregated funds, hard transaction and daily limits, approved counterparties, fast revocation, and a tested kill switch. An independent system checks the rules before signing.
The tools feeding agents create another risk. Snyk scanned 3,984 public agent skills in February and found at least one security issue in 36.82%. It confirmed 76 malicious payloads involving credential theft, backdoors, or data exfiltration.
Gaur sees prompt injection as the dominant pattern: “Prompt injection is the dominant pattern: an agent takes instruction from untrusted content it was asked to read and executes it as though the principal had asked.”
A defensible audit trail therefore needs the agent identity, signed mandate, policy version, transaction, source data, and any approved exception, written when payment occurs. The chain provides one part.
Gaur’s standard is shorter: “Provable, revocable and bounded.”
Without those properties, companies are left with an immutable receipt for a decision they cannot defend.
The post Who is Responsible When an AI Agent Loses Your Money? appeared first on BeInCrypto.
-
Fashion6 days agoWeekend Open Thread: Madewell – Corporette.com
-
Crypto World2 days agoSpaceX stock could rise 75% to $240, JPMorgan says
-
Business5 days agoMusk’s Tesla, SpaceX Confirm $16.8 Billion ‘Terafab’ Chip Plant as World’s Largest Building in Texas
-
Crypto World6 days agoanatomy of crypto’s biggest liquidation event since 2021
-
Crypto World3 days agoA $30 Billion AI Fund Implodes, Now the SEC Is Investigating Wall Street’s Role
-
Politics5 days ago6 months on, Irish renters crushed by effects of government housing bill
-
Crypto World2 days agoDid Trump Just Move SpaceX Stock With One Truth Social Post?
-
Business5 days agoMystery AI Model ‘Ox Alpha’ Draws Developers With Free Access as Chinese Lab Origins Remain Debated
-
NewsBeat6 days agoThe ‘Lucky Dip Gang’ causing carnage for clicks: After five thugs were killed speeding in the wrong direction on a motorway, GUY ADAMS investigates a sick new trend… and why police aren’t even allowed to pursue them
-
Crypto World6 days agoNvidia Stock Suffers Longest Losing Streak Since 2022: Will Q2 Earnings End It?
-
Business3 days agoModerna CEO warns China is pouring state money into mRNA technology
-
Business3 days agoNVIDIA Stock Drops Nearly 2 Percent to $210 on Seventh Losing Day Ahead of Critical AI Earnings
-
Tech6 days agoUnitree’s New Superman Robot Claims to Outjump and Outrun Every Human, Usain Bolt Included
-
Business6 days agoUK firms in critical financial distress rise 9% to 53,756
-
Sports5 days agoDeshaun Watson fires back at Browns fans after being booed: ‘It’s a disrespectful thing’
-
Crypto World5 days agoGoogle Gemini AI Predicts Ethereum Could Become the Trade Retail Misses in 2026
-
Tech6 days agoFaster chip in a familiar form factor
-
Crypto World7 days agoSEC Regulation Crypto vs CLARITY Act: which framework wins
-
Business6 days agoWill Tesla Stock Be Higher or Lower a Year From Now? Here’s What Wall Street Analysts Are Saying
-
Business4 days agoTesla Recalls Nearly 3 Million Cars in China Over Hidden Door Handles Linked to Multiple Deaths Amid New Ban

You must be logged in to post a comment Login