Last week I was lucky enough to be in Paris for the Proof of Talk conference, held at the Louvre. It was a chance to sit down with three senior figures leading the charge toward digital cash and a new form of institutional money: Emma Landriault, Global Executive Director of JPM Coin at JP Morgan; Tero Reuna, CEO of Paxos EU; and David Cunningham, Head of Institutional at Consensys. Thanks to our friends at X Ventures for putting on a beautifully curated event in a very special venue.
We covered a lot of ground, but two tensions stood out for me.
Stablecoins vs. Deposit Tokens: A Live Clash
The regulatory picture has shifted fast. MiCA is law across Europe. The GENIUS Act, passed in the U.S. in 2025, mainstreamed stablecoins. But it also arguably hamstrung stablecoin issuers from paying yield and attracting more customers. It seemed to advantage banks issuing their own “deposit tokens” over stablecoin issuers coming into the market.
Against that backdrop, JPMorgan launched JPMD on Base last November, positioning deposit tokens (which are tokens representing real deposits) as superior to stablecoins for institutional use: same programmability, but with balance sheet treatment and yield eligibility.
Emma made the case with characteristic precision. What JPMD offers that USDC cannot is the full weight of a regulated balance sheet behind every token; it's not a claim on a reserve that may or may not exist. It's a deposit on a ledger you can trust.
Tero pushed back from the stablecoin side. Regulated stablecoins in 2026 are not the Wild West instruments of five years ago: they are regulated, audited, and increasingly the default rails for cross-border payments and settlement. The ceiling for deposit tokens, he argued, is interoperability. They work beautifully within a bank's own ecosystem and much less cleanly outside it.
The honest answer, which emerged as the panel developed, is that these instruments probably serve different customers rather than competing head-to-head. Deposit tokens for large institutional wholesale flows. Stablecoins for open-network payments and programmable settlement.
Canton Privacy vs. Protocol-Level Privacy
As a moderator who works a privacy blockchain, I wanted to know where the panelists stood on privacy.
Emma diplomatically praised the Canton Network, the permissioned blockchain ecosystem that allows institutions to transact in a walled garden with selective disclosure (JP Morgan is a Canton member and plans for JPM Coin to trade natively on the network). It's a reasonable solution to a real problem: large financial institutions need privacy from their counterparties and from the market, and they need it in a form their compliance teams can defend.
But David, coming from the Ethereum world, made a point that I think deserves more attention. There is real space in the market for protocol-level privacy – ZK proof-based privacy built into public infrastructure – that doesn't require you to trust the walls of the garden.
Canton privacy is permissioned privacy: you're private because the system administrators decide who can see what. Protocol-level privacy is mathematical privacy: you're private because the cryptography makes it impossible for anyone to see what they're not supposed to see, including the infrastructure operators.
David, who is now leading institutional work at Consensys, a key Ethereum developer, was previously head of digital asset partnerships at Citi, so he’s uniquely positioned to see across the blockchain-financial institution divide.
His distinction matters more as these systems scale. Permissioned privacy concentrates trust. Protocol-level privacy distributes it. For institutions that are genuinely worried about counterparty surveillance – not just regulatory compliance – the long-term architecture looks more like the latter.
The panel didn't resolve anything. But it sharpened the debate for me between two versions of privacy. The institutions that figure that out earliest where they stand will have a structural advantage.
Privacy Roundup
Security researcher Taylor Hornby discovered a critical counterfeiting vulnerability in Zcash's Orchard shielded pool – an under-constrained circuit element that could have allowed unlimited, undetectable counterfeit ZEC. The bug has been fixed.
AnomaPay V1 went live on BNB Chain and Ethereum, introducing multichain support with ZK-powered private transfers in 15 seconds, passkey onboarding, and shareable payment links – no wallet addresses needed.
Veil shipped MCP 0.2.0, letting AI agents make private x402 payments on Base directly from the Veil shielded USDC pool.
Labrys and Aleo previewed Confidential Sponsor at the Booz Allen × a16z event in DC, a private payments architecture for crisis response that keeps sender, receiver, and fee-payer data dark, denying adversaries the intelligence public chains leak by default.
Latest on Miden
Last week, we announced the first batch of Guardian operators: Gateway, LambdaClass, OpenZeppelin, Qash and Sempo.
The last two, Qash and Sempo, are neobanks building on Miden. Qash aims to solve onchain payroll for blockchain teams and Sempo is building borderless finance for the MENA region.
.jpg)
Thanks for reading this edition of Privacy Dispatch. Please subscribe to receive this newsletter every Tuesday.
Till next time.