TOKEN2049 Singapore Highlights Stablecoin Payments, Tokenisation and AI
The 7–8 October conference at Marina Bay Sands puts payment distribution, institutional settlement and AI trading on the programme, with Circle running a parallel Arc and USDC showcase.
By Daniel Osei, Journalist, Payments and Cards
TOKEN2049 Singapore runs at Marina Bay Sands on 7–8 October 2026, bringing stablecoin payments, tokenised markets and AI trading into the same conference programme. For payments and financial-infrastructure readers, the useful highlights are the sessions on distribution and settlement, alongside Circle’s two-day Arc and USDC showcase.
This is a programme-based highlights report. The official agenda establishes the listed sessions and speakers; it does not establish what each participant said on stage or prove that a product demonstrated at the event is generally available.
Stablecoins meet the payment distribution business
The October 8 main-stage programme lists a payments panel with May Zabaneh of PayPal, Stephen Richardson of Fireblocks, Ivan Soto-Wright of MoonPay and Farooq Malik of Rain. The mix brings merchant-facing payments, institutional infrastructure, crypto distribution and card infrastructure into one discussion.
That is a useful commercial frame for stablecoins. Issuing a token and moving it on a blockchain are only part of a payment flow. A usable payment product also needs customer access, compliance checks, merchant acceptance, conversion and local payout arrangements. Those operational questions provide a more useful way to assess announcements around the conference than a headline transaction-speed claim.
The October 7 programme also lists a stablecoin-payments session with World Liberty Financial’s Zach Witkoff and Zak Folkman. The appearance of executives on an agenda is evidence of the conference’s subject matter, rather than independent evidence of adoption or regulatory approval.
Institutional tokenisation is about liquidity and settlement
The October 8 opening institutional-market panel lists Franklin Templeton’s Jenny Johnson, Binance’s Richard Teng and Canton’s Yuval Rooz. Its focus is tokenised assets, global liquidity and onchain settlement. On October 7, the programme also lists Alpaca’s Yoshi Yokokawa for a session on the infrastructure missing from tokenised-stock markets, and Superstate’s Jim Hiltner for a discussion of distribution bottlenecks.
The distinction matters: creating an onchain representation of an asset does not, on its own, provide a liquid market or settle the cash leg. For a financial institution, the relevant questions are ownership rights, custody, who can trade, the settlement asset and the route back into conventional financial systems. The agenda shows those questions receiving dedicated attention; it does not demonstrate that the participating companies have solved them through a new joint deployment.
Circle brings Arc and USDC programming to Marina Bay Sands
Circle’s own event listing describes Circle House on October 7–8 at Marina Bay Sands, Level 3, Angsana 3A, with programming scheduled from 9:30am to 5pm Singapore time each day. Circle advertised Arc and USDC demonstrations and sessions with infrastructure partners. Its listing says entry required registration but did not require a TOKEN2049 conference pass.
That programme gives developers and payment businesses a separate venue to inspect the infrastructure alongside the conference. The listing does not specify a new launch, transaction volume or completed customer deployment. Demonstrations should be evaluated against product documentation and availability conditions before being described as production payment services.
AI trading is visible, but autonomous payments need separate evidence
The October 7 programme lists Post Fiat’s Alexander Good on blockchain-based AI trading, Nansen’s Alex Svanevik on agentic trading and NEAR’s Illia Polosukhin on privacy in AI and finance. These are programme highlights, without attributed quotations or claims about session outcomes.
Trading agents and payment agents raise different operating questions. For an enterprise payment, authorization, spending limits, access to funds and reconciliation matter alongside the model’s ability to choose an action. A conference demonstration or AI-themed presentation is therefore not sufficient evidence that an agent can independently make a regulated payment in a live customer workflow.