If you were not in the room on 2–3 June 2026, the public record of what was argued at Proof of Talk in Paris is thinner than the search results make it look. A dozen pages appear to cover the edition. Read them and most collapse into two organiser press releases, one issued before the edition and one after, syndicated in near-identical form across FXStreet, The Block and Metaverse Post among others. Not one of those pages carries a sentence in quotation marks attributed to a named speaker. Neither does the closest thing to an independent write-up, Crypto Briefing's 10 June piece, which names panels and institutions and quotes nobody.
This piece sets out what is genuinely on the record from the archived Paris edition, organised by argument rather than by speaker, with each claim traced to whoever published it. Where the record is thin, we say so. Where a repeated claim does not survive checking, including several that originated with us, we say that too.
What the record actually contains
There are two layers, and they are not equal.
The first is the published programme: session titles, times, stages, moderators and panel line-ups. That is a primary document. It tells you what the organisers believed the live arguments were, who they put on each side of them, and which working journalist held the chair.
The second is the organiser's own material, a pre-event announcement in May and a post-event wrap in June listing announcements made across the two days. It is a fair source for what a company chose to announce and a poor one for what anyone argued, because it paraphrases throughout and quotes nobody. The syndications reproduce it without doing that work.
There is no third layer. No transcript has been published, the session recordings are not yet online, and nothing below paraphrases remarks that no outlet printed. Where we describe a disagreement, we are describing how the programme was constructed, not what was said inside it. The alternative is inventing quotations, which several summaries of this event have effectively already done.
Two points of fact are worth fixing, since sources vary. The edition ran at the Musée des Arts Décoratifs in the Louvre Palace, not in the museum galleries, and it ran across four rooms: the Hecto Main Stage, the Taostats Stage, the X Ventures Masterclass Stage and the Roundtable Lounge. La Nef at the Musée des Arts Décoratifs, which some accounts have taken for the conference floor, was where the evening drinks were held on both days. Capacity was 2,500 and capped, admission was by application and review, and Paris has sold out in 2024, 2025 and 2026.
The tokenisation case, and a report published the day before
Tokenisation took more main stage time than any other subject. Its clearest expression was Day 1, 11:45–12:25, "The Great Crossover: Asset Management for the Onchain Investor", pairing Franklin Templeton chief executive Jenny Johnson with Blockstream co-founder Adam Back under Eleanor Terrett as moderator. The theme ran through Day 2 in "Scaling the Distribution Layer for Tokenised Markets", with Robinhood's Johann Kerbrat, BlackRock's Nikhil Sharma, Aave Labs founder Stani Kulechov and Tom Zschach, moderated by Bloomberg ETF analyst Henry Jim, and again that afternoon in "The Tokenisation of Everything: RWAs, Institutions, and the Next Market Structure", with BNP Paribas, Ava Labs, Raptor Digital and BitMart on the panel under Cointelegraph chief executive Yana Prikhodchenko. Johnson returned later on Day 1, at 15:35, with MoonPay Institutional's Caroline D. Pham, for a fireside on embedding digitally native money market funds directly into on-chain transactions. The organiser's release credits a Franklin Templeton and MoonPay partnership to that stage.
The substantive document behind all of it was Citi's Global Perspectives & Solutions report, Tokenization 2030: Wall Street On-Chain. Ronit Ghose, whom The Defiant identifies as Citi's global head of Future of Finance and whom our own programme lists only as a senior executive at Citi, walked the room through it on Day 1, 15:00–15:30, in "What Institutions Need From Blockchain Infrastructure" alongside Mysten Labs co-founder Evan Cheng, moderated by GSR's Frank Chaparro. The projections are Citi's, not ours, and they are set out on Citi's own page for the report, which is where anyone quoting them should go. The Defiant, one of the few outlets to write the report up independently, reported that Ghose was scheduled to present it on the main stage, citing organisers.
Our own release says the report was unveiled publicly for the first time anywhere in the world from the Proof of Talk stage. That is wrong, and the evidence against it is partly ours. Citi dates publication to 1 June, the day before Ghose took the stage, and our own account said as much at the time, posting that Citi had just published the report and that Ghose would walk the room through it the following day. A report presented the day after publication, by its author, to a room of allocators is still a substantial moment. It is not a world premiere, and we would rather correct it than let it stand.
The resistance was in the same room
Programming Jenny Johnson opposite Adam Back is the closest thing the edition had to a formal debate: an asset manager whose firm has built tokenised fund infrastructure, against a Bitcoin infrastructure founder sceptical about putting everything else on a chain. What either of them said has not been published anywhere we can find, so we will not characterise it.
What the programme does show is that the sceptical case was given room, and that it was specific rather than abstract. "After the Flush: What October 10th Revealed About Crypto Market Fragility" took the 13:45 slot on Day 1 and put Hamilton Lane's Victor Jung, Upbit Global chief executive Alex Kim, BitMEX chief executive Stephan Lutz and STBL's Avtar Sehra on a panel premised on structural fragility rather than adoption. "Institutional DeFi: The Use Cases Worth the Risk", with Societe Generale-FORGE's Stéphanie Cabossioras, Perkins Coie's Dax Hansen and FGNexus chief executive Maja Vujinovic, concedes the risk in its title and argues about which uses clear it. "DeFi Under Attack: Securing Onchain Finance and the Machine Economy" brought in Ledger's chief technology officer Charles Guillemet and Mysten Labs' chief cryptographer Kostas Chalkias.
An edition that only wanted the bullish case would not have given the fragility panel a prime Day 1 slot.
Stablecoins, deposits and the rails underneath them
Stablecoins were the second spine of the programme, and the sessions split into two arguments. One treated stablecoins as an investable sector: "Investment Opportunities in the Rise of Stablecoins" on Day 1, with Haun Ventures general partner Diogo Mónica and Dragonfly's Rob Hadick, moderated by Serotonin's Amanda Cassatt. The other was about stablecoins as plumbing, and it belonged to the incumbents. "The New Architecture of Digital Cash: Stablecoins, Tokenised Deposits and Money Market Funds" put JP Morgan's Emma Landriault, who runs JPM Coin, next to Consensys and Paxos. "Where Stablecoins Meet Credit" added Coinbase Asset Management, Citi's head of investor services and Fasanara.
A tokenised deposit, a money market fund share and a stablecoin are three different legal instruments competing to be the same thing, which is settlement cash that earns. Putting a bank, a fund manager and an issuer on one panel is a statement about who thinks they will win.
The one on-stage line we can quote from any Proof of Talk edition sits inside this argument, and it is from 2025, not 2026. As Cointelegraph reported from the 2025 panel "Stablecoins: Programmable Money in a Digital World", Mónica said of holding a tokenised dollar that "it's actually much better than having a dollar in a commercial bank". He returned to the Paris stage in 2026 on a different panel, and there is no published record of what he said there. The quote belongs to 2025, which is more than the circulating summaries acknowledge.
Per the organiser's release, a Stablecoin Roundtable convened senior figures including representatives of Aave, the Bank of England, BlackRock, Paxos and Robinhood. That is an organiser claim, not reporting.
Agentic commerce arrived as an assumption
The striking thing about the AI programming is that nobody was asked to defend the premise. After the opening speech at 10:20, Day 1 began in earnest at 10:30 with "Beyond Stablecoins: Mastercard and the Future Rails of Agentic Commerce", a session with Mastercard's Ken Moore moderated by Christine Lee. Fundstrat co-founder Tom Lee took a solo slot at 12:30, titled "Blockchains Are Central to the Future of Finance and Agentic AI". Bittensor co-founder Jacob Steeves took the same stage at 10:05 on Day 2 with "Decentralising Intelligence: The Unification of AI and Bitcoin", moderated by Mark Jeffrey.
Read the titles together and a disagreement surfaces that no outlet noted. Mastercard's framing is that agentic commerce needs rails, and that an incumbent network builds them. The Bittensor framing starts from the opposite end. The agenda's description of Taostats founder Marcus Graichen's talk "The Thin Line" argues that the agent economy needs both a permissionless protocol and an accountable industry built on top of it. That is the sharpest statement of an argument to come out of the edition, and it comes from our own agenda rather than from coverage. SimplyTao's track recap confirms the sessions ran and paraphrases them, but quotes no speaker on the substance of an argument.
Custody, access and settlement
The market structure question came most directly on Day 1 at 17:20, in "Banks, Exchanges, and Asset Managers", with Morgan Stanley's head of digital assets Amy Oldenburg, Upbit Global's Alex Kim and VanEck's Matthew Sigel, again under Frank Chaparro. Settlement infrastructure ran through two sessions featuring Jorgen Ouaknine, "Building Canton: A Conversation with the Foundation Board" on Day 1 and "Building DLT Infrastructure for Next Generation Capital Markets" on Day 2, the latter moderated by The Block's Gareth Jenkinson.
Two caveats belong here, and both cost us something. Our agenda lists Ouaknine as co-chairman of the Canton Foundation; our speaker directory lists him as global head of innovation and digital assets at Euroclear. Both may be true at once, and we will not pick one silently. Separately, the release credits him with confirming on stage that Euroclear's Pythagore tokenisation project, covering France's short-term commercial paper market, is on track for the fourth quarter of 2026. Pythagore was announced with the Banque de France on 10 October 2025, so what happened in Paris was a progress update rather than news, and the Crypto Briefing analysis that circulated alongside the claim sets out Ouaknine's views without mentioning this event at all.
The release also states that Zodia Custody chief executive Julian Sawyer confirmed a sale of the business to a traditional finance entity. Sawyer was on the programme, at 10:00 on Day 2, but on the Taostats Stage rather than the main stage, on a panel called "How Institutions Actually Access Tokenised Assets" moderated by Vancelian's Xavier Gomez. Nothing published places the announcement in that room, so the claim stays where we found it, in the release.
What we are not repeating
The claim that Bloomberg, Fortune and Forbes covered the edition rests solely on our own release, and we could not locate an article or a segment from any of them.
The counting is inconsistent too. Our pre-event release gives one speaker total in its headline and a larger one in its body, and the live speaker directory lists a third, larger still. Our own site renders the capital in the room as €18T where the press material used dollars, and a headline number should not change currency between two properties owned by the same company. On one point the sources agree: our release and Crypto Briefing both call Paris 2026 the fourth edition.
The through-line
Across five clusters of sessions, one question repeats. Not whether assets move on-chain, but who operates the layer between the asset and the client. That is what the Franklin Templeton and Blockstream pairing was about, what the distribution and digital cash panels were about, and what a session on who controls institutional access is about by construction. Every panel that looked like a technology panel was a distribution panel underneath.
The second through-line is the machine economy, which entered the programme as a shared assumption and left it as a contested one, the disagreement sitting between an incumbent network's rails and a permissionless protocol. The third is that the resistance was never abstract. Nobody was booked to argue that tokenisation will not happen. Several people were booked to argue that a particular structure is fragile, or that a particular use case is not worth its risk, which is what institutional scepticism actually looks like.
The source of record
Everything above is what can be sourced today, and it is not a substitute for the sessions themselves. Full recordings are the source of record. Past editions sit on the official Proof of Talk YouTube channel, and the Paris 2026 recordings will go up on the same channel. We are not giving a date.
Until then, the agenda and the session pages are the archive of who argued what and when. That roster is closed and bound to 2–3 June 2026. The next editions are the inaugural Gulf edition at Louvre Abu Dhabi, Saadiyat Cultural District, on 3–4 December 2026, capacity 2,000 and capped, and the Paris flagship at the Musée des Arts Décoratifs on 2–3 June 2027. Both are by application and review, and no speaking slot is ever sold.