Today on The Onchain Dispatch: the L2 shakeout keeps producing winners and zombies, a16z drops $2.2B into a new crypto fund, AI-agent micropayments inch closer to a real media business model, and Ohio quietly becomes the first state to accept crypto for agency payments. Plus a Solana upgrade story Anza would rather you not read.
Trump's May 19 Executive Order 14405 was officially published in the Federal Register on May 22, formalizing the 120-day Fed review of payment-account access for digital asset and non-bank fintech firms, plus 90-day directives to six federal regulators (SEC, CFTC, FDIC, OCC, NCUA, CFPB) to streamline fintech rules. SEC Commissioner Hester Peirce separately clarified that the forthcoming tokenized-equity innovation exemption will be narrowly tailored — applying only to digital representations of existing secondary-market stocks, not synthetic assets.
Why it matters
The Federal Register publication starts the actual regulatory clock. Combined with Peirce's narrow-tailoring guidance, the two-rail tokenization architecture (DTCC institutional rail + crypto-native exempted rail) is now the operating reality through summer. For BD and content planning, the dates to mark are: 90 days from May 22 for streamlined fintech rules across six agencies, 120 days from May 22 for the Fed's master-account framework. Both arrive before year-end and before the Senate floor vote on CLARITY.
ACDE #237 surfaced an open developer dispute over EIP-8037, which would sharply raise upfront gas costs for contract deployment and storage to address Ethereum's ~197 GiB/year state growth (currently ~390 GiB, projected to hit the 650 GiB node-decentralization threshold in under 1.6 years). Vitalik rejected proposed workarounds using client-side storage and cryptographic proofs as too complex. In parallel, Vitalik publicly called for simpler node architecture (Docker wrappers, unified beacon+execution clients à la Nimbus).
Why it matters
The Glamsterdam scope was finalized last week; what's new is the substantive disagreement inside the protocol cluster about the economic-incentive layer of state management. This is the next governance flashpoint after the EF departures — and unlike the Feist proposal, it's a fight that has to be resolved in code, not on X. For educators, the framing 'devs pay once, nodes pay forever' is the clearest one-line explanation of Ethereum's state problem that's emerged in months. Worth banking for ETHDenver and UETH curricula.
Former a16z Crypto investor Michael Blau is building Drip, a platform where AI agents pay newsletter writers and Substack creators micropayments per article, settled on-chain. The launch target is finance and tech analysis writers selling content directly to autonomous AI consumers rather than human subscribers.
Why it matters
This is the first product-level test of the agent-payments thesis aimed squarely at independent publishers, and it lands the same week Cloudflare's Matthew Prince was on Bankless proposing the same architecture (stablecoins + HTTP 402 + fast settlement) as the only viable answer to AI's 60,000x-collapse of OpenAI referral traffic. For an operator running a Web3 media company, this is the closest thing yet to a working answer to 'how does premium analytical content get paid for when humans aren't the primary readers.' Worth tracking as both a competitor and a potential rail for your own distribution.
The Bankless situation has now moved from internal crisis to external case study. New since yesterday: David Hoffman publicly confirmed he sold the last of his ETH, and creator-economy media (CORQ, American Bazaar) are now using Bankless as the cautionary example in a broader piece on creator-journalism transitions — meaning the story has escaped the crypto press. Press Gazette counted 3,434 journalist redundancies across UK/US in 2025, with named journalists like Jim Waterson and Carole Cadwalladr moving direct-to-Substack.
Why it matters
Hoffman selling his ETH while taking over an Ethereum-focused media brand is a new and concrete positioning problem that wasn't in yesterday's coverage — it's the detail that will make this story harder to contain. The second development is that the Bankless narrative has now crossed into mainstream creator-economy coverage, which means Web3 media operators are being watched by peers outside the ecosystem. The editorial vacuum this creates is the real opportunity signal.
Andreessen Horowitz announced Crypto Fund 5, a $2.2 billion vehicle to deploy over a decade across stablecoins, perps, lending, prediction markets, and tokenization — explicitly named focus areas. CTO Eddy Lazzarin was elevated to general partner. The fund is materially larger than recent peers (Haun $1B, Dragonfly $650M) and lands as generalist VCs continue rotating to AI.
Why it matters
Two signals worth holding side-by-side. First, the explicit thesis: a16z is not funding more rollup-as-a-service or generic infra — it's funding the categories where revenue is already on-chain (stablecoins at $320B, perps, tokenized treasuries). That validates the 'revenue chain' framing now hardening across Bitwise, JPMorgan, and Hunter Horsley's writing. Second, the timing: this is the largest crypto-dedicated fund announced post-CLARITY-Act markup and post-Trump fintech EO, meaning a16z is pricing in a regulatory environment that didn't exist 18 months ago. For BD operators, the addressable categories just got named in writing.
On-chain tokenized fund market cap reached $32.4B with Ethereum capturing 59.6% share. A parallel Finance Feeds analysis surfaces what changed: Circle's USYC ($2.2B) overtook BlackRock's BUIDL ($2.4B) in March not on fund quality but because USYC integrated as off-exchange collateral on Binance via BNB Chain. FINRA cleared Securitize Markets on May 4 as the first US broker-dealer permitted to custody tokenized securities and settle atomically against stablecoins.
Why it matters
The framing matters here: tokenized treasuries are no longer a custody story or a yield story — they are a composability story, where integration breadth across exchanges and lending markets determines flows. That has direct BD implications: if you're a media or education partner, the right relationships are with the integration layer (Securitize, Ondo distribution, Binance off-exchange settlement), not the issuer brands. The Securitize-FINRA clearance is the first regulated template, and it will be copied.
YZi Labs (formerly Binance Labs, $10B+ AUM) opened applications for its 10-week EASY Residency Season 4, hosted on-site in Gelephu Mindfulness City — Bhutan's new Special Administrative Region. Up to $500K per founder, full accommodation, meals, and workspace covered. Priority sectors: global payments infrastructure, agentic economy, DeFi, AI-bio. Bhutan signed a Double Taxation Avoidance Agreement with Singapore on May 12 to support the SAR's tax positioning.
Why it matters
GMC is the most concrete sovereign-jurisdiction play in Web3 since Próspera, and YZi committing a flagship cohort there moves it from 'rumor on Crypto Twitter' to operational reality. For BD operators tracking where founder talent will physically concentrate over the next 18 months, this is a Tier 1 signal. Worth a placement conversation: international founders looking for English-language media coverage of GMC have very few options yet.
Variational closed a $50M Series A led by Dragonfly Capital with Bain Capital Crypto and Coinbase Ventures, bringing total to ~$60M. The Arbitrum-based DEX launched RWA perpetuals on gold, silver, copper, and oil alongside crypto pairs. Separately, Foresight Ventures led $30M into PopDEX, and Solstice Finance crossed $400M TVL on Solana with NYSE-listed Bullish joining its institutional allocator base.
Why it matters
Three rounds, one thesis: institutional capital is no longer betting on generic perp DEX infra — it's betting on differentiated execution venues with explicit institutional pipelines. Dragonfly + Bain + Coinbase Ventures is the most institutional combination a perp DEX has assembled, and the RWA perpetuals angle (commodities, not just crypto) tracks directly with the broader 'revenue chain' framing. Worth pulling these three rounds together as a data point in any commentary on where 2026 capital is actually flowing.
DeFi distribution protocol Turtle migrated $5.5B in TVL from LayerZero to Chainlink CCIP, joining Solv, Lombard, Huma Finance, and Kraken in similar migrations — over $4B in completed or initiated moves. This is the largest single migration yet triggered by the $293M KelpDAO exploit (May 18), which you've already seen traced to a 1/1 DVN single-validator LayerZero setup. The new detail: Turtle's $5.5B dwarfs earlier migrations and pushes the total migration volume past $9B, cementing this as a category-level repricing rather than a reaction by a few cautious protocols.
Why it matters
The KelpDAO exploit and the migration wave have been in your feed since May 21. What's new here is the scale: Turtle alone is larger than all prior migrations combined. That changes the story from 'some protocols are switching' to 'LayerZero's configurable DVN model has triggered a structural flight.' The competitive framing has also sharpened — Chainlink's 16-node decentralized verification model is now being marketed as the safe-harbor alternative, and Wormhole/Axelar's silence is becoming conspicuous.
Solana's Alpenglow consensus upgrade — pitched as the largest in network history and approved by 98% of validators — broke during May 11 testnet migration with bugs in TowerBFT and proof-of-history. Lead dev team Anza disclosed the failures only on a small validator call; mainstream press cycle and official channels described the launch as successful.
Why it matters
Compare and contrast: Ethereum's ACDE #237 published the full Glamsterdam interop punch-list including which EIPs got pushed and why. Anza chose to keep Alpenglow failures off the public record while celebrating the milestone. For an Ethereum educator, this is rare ammunition — not because Solana having bugs is news (every L1 does), but because the asymmetry in disclosure norms is a clean teaching example of how governance transparency actually varies across ecosystems. Useful raw material for commentary on developer trust.
Ohio Treasurer Robert Sprague launched Buckeye Billfold on May 21, allowing residents and businesses to pay state agencies — including the BMV and courts — using cryptocurrency, credit card, or bank transfer. Crypto is auto-converted to USD at transaction time via vendor Grant Street Group. Ohio claims it is the first state to authorize statewide crypto payments to government.
Why it matters
This is the operational template South Carolina, Minnesota, and Wyoming have been gesturing toward — concrete vendor relationships, conversion mechanics, and a list of participating agencies. Unlike a strategic reserve bill, this is something other state treasurers can copy verbatim. The choice to use a private payment processor rather than a state-run rail also tells you what 'crypto-friendly state policy' actually looks like in production: integration plumbing, not sovereign infrastructure. Watch for which states announce similar pilots within 60 days.
Helium Mobile announced on May 22 that its free Zero Plan ends June 11; existing subscribers — many grandfathered after the first sunset — get auto-migrated to a $15/month plan unless they opt out. The carrier has also been deleting negative comments and banning users on its subreddit. Separate Solana data shows Helium Mobile alone accounts for two-thirds of Solana's $22M+ cumulative DePIN revenue.
Why it matters
This is the moment of truth for the DePIN incentive model. Helium Mobile is the most-cited proof point that token-subsidized wireless can compete with carriers; pulling the free tier reveals whether real consumer demand exists once the subsidy stops or whether the subscriber base was always a token-emissions artifact. Watch the churn data — and watch the community moderation pattern, because the subreddit bans are a governance signal that matters as much as the pricing. Either Helium emerges as the DePIN survivor case study or as the cautionary tale.
El Salvador's Bitcoin Office announced its fourth CUBO+ developer cohort will graduate June 21, bringing the total to over 1,000 Salvadorans trained in Bitcoin and Lightning development. The curriculum combines technical skills with what the Office calls 'mindset' work, built in partnership with weSpark_io and a refreshed Bitcoin Diploma. The pipeline is structured to feed both local builders and the Lightning developer market globally.
Why it matters
Most national-level Web3 education programs are PR exercises; CUBO+ has produced four cohorts and a measurable graduate count, which makes it a more credible reference than most US university blockchain programs. For an Ethereum educator, the comparison worth making is that El Salvador's pipeline is producing real Bitcoin/Lightning specialists at scale while Ethereum-focused programs (Encode, UETH, university chapters) still struggle with continuity. There's a content angle here on what the EF cluster could learn from a smaller, more focused state-level program.
Cloudflare CEO Matthew Prince, on Bankless, laid out the most concrete framing yet of the AI-publisher collapse: Google referral traffic down 20x since AI Overviews became default, OpenAI down 1,500x, Anthropic down 60,000x. His proposal: stablecoins + HTTP 402 + ultra-fast blockchain settlement (target 100M TPS) to charge AI agents micropayments per crawl, with revenue pooled in a Spotify-style model. Existing chains (Tempo, Arc, Base) currently max at ~2M TPS — a real infrastructure gap.
Why it matters
Prince is the most credible non-crypto-native voice to publicly argue that Web3 payment rails solve the AI economy's most concrete problem. The framing — 'ad-supported internet collapses by H1 2027 unless agents pay' — is a clean story arc that bypasses the usual crypto talking points and arrives at the same conclusion. It also surfaces a real infrastructure spec gap (100M TPS) that nobody in Web3 is currently shipping toward. Good raw material for commentary, and a useful counterweight to the 'crypto solves nothing real' frame.
The L2 shakeout is now a category event, not a series of one-offs Zero Network, Everclear, and Syndicate Labs shutting down on the same day was last week's news; this week the second-order effects show up — Turtle moving $5.5B from LayerZero to Chainlink CCIP, Aztec topping GitHub activity rankings, and analysts openly using the phrase 'zombie chain.' The consolidation has stopped being a thesis and started being the operating environment.
Capital is rotating from infrastructure platforms to specialized revenue chains a16z's $2.2B Crypto Fund 5 explicitly targets payments, DeFi primitives, and tokenization — not generic L2 toolkits. Variational, Solstice, and PopDEX are pulling institutional money for vertical-specific perp and yield infrastructure. Hyperliquid's FDV briefly crossing Solana's last week reinforces the framing: the market is paying for fee-generating specialization, not modular optionality.
States are quietly building the real crypto policy stack while DC argues over CLARITY Ohio launched Buckeye Billfold for statewide agency crypto payments, Minnesota authorized state-bank custody, South Carolina banned CBDCs and protected self-custody. None of these waited for federal market-structure legislation. The sub-federal layer is where the actual operational templates are being written.
Stablecoins-as-distribution-rail is the new media-and-payments thesis Drip's AI-agent micropayments, Kura's USDC merchant network in Haiti, Sorted's $4.4M for feature-phone wallets, AirAsia MOVE's Solana tenge integration, Deel's Polygon payroll, Cloudflare's HTTP 402 proposal — every one of these treats stablecoin settlement as the layer that makes a non-crypto product work. The frame has shifted from 'stablecoins as crypto' to 'stablecoins as plumbing.'
The EF crisis is producing competing institutional designs, not just departures Dankrad Feist's $1B counter-organization proposal got formal write-ups across ForkLog, Unchained, and Futunn this week — moving from X discourse to actual governance proposal. Meanwhile ACDE #237 confirms the protocol cluster is still shipping (Glamsterdam, EIP-7928, Hegota), and Vitalik is publicly pushing simplified node architecture. The institutional question is now whether one EF or two complementary orgs run Ethereum's next decade.
What to Expect
2026-05-29—SFI / Bank for International Settlements Master Class on blockchain in finance — a useful read on how central banks are framing institutional blockchain literacy.
2026-06-05—Trident Digital Tech Holdings launches Ghana tax digitization platform with planned RLUSD integration; targets 530,000 MSMEs in first 12 months.
2026-06-11—Helium Mobile discontinues its free Zero Plan; existing users migrate to $15/mo. A real test of whether DePIN can transition from subsidy to revenue.
2026-06-21—El Salvador Bitcoin Office graduates fourth CUBO+ developer cohort — 1,000+ Salvadorans through the Lightning developer pipeline.
2026-09-09—Web3 Warsaw 2026 — Central Europe's largest blockchain conference; potential BD and editorial venue for ecosystem partnerships.
How We Built This Briefing
Every story, researched.
Every story verified across multiple sources before publication.
🔍
Scanned
Across multiple search engines and news databases
919
📖
Read in full
Every article opened, read, and evaluated
188
⭐
Published today
Ranked by importance and verified across sources
14
— The Onchain Dispatch
🎙 Listen as a podcast
Subscribe in your favorite podcast app to get each new briefing delivered automatically as audio.
Apple Podcasts
Library tab → ••• menu → Follow a Show by URL → paste