The Central Bank's timeline for 2026 stopped carrying the word "CBDC". Drex, announced in August 2023 as "the most innovative central bank digital currency in the world", will debut in the second half of this year as a centralized lien reconciliation infrastructure — a banking back-office tool to verify whether assets pledged as collateral in credit operations are already linked to other obligations.
No blockchain. No tokenization. No smart contracts. No public access. No end-user.
What remains of the project that Roberto Campos Neto presented as "the beginning of banking system tokenization" fits in a single paragraph in the operational plan: financial institutions consult, via the Central Bank's centralized API, whether a given asset already has a lien in another registry. When it does, the system blocks the duplication. When it doesn't, it clears the operation. End of scope.
This article closes the arc we opened when B3 announced its real-backed stablecoin and buried Drex as a popular CBDC. At that moment, the signal was already clear: the tokenization gap for the real would be filled by the regulated private sector, not by the Central Bank. Today, with confirmation of Drex 2026's narrow scope, the burial is official. And there is a third character in this story who warned since 2024: the Federal Court of Accounts.
2023: the vanguard that was announced
When it launched Drex in August 2023, the Central Bank under Campos Neto set an ambitious narrative. It was Latin America's first retail CBDC, based on Hyperledger Besu — an Ethereum fork on a permissioned layer — with smart contract modules planned for automated settlement, tokenized accounts and direct access via citizen wallets.
The public pitch combined five promises:
- Financial inclusion: direct access to banking services through the digital layer, without traditional intermediation;
- Native asset tokenization: real estate, vehicles, public securities circulating on-chain;
- Atomized settlement: smart contracts for Delivery vs. Payment (DvP) in seconds;
- Social benefit payments: Bolsa Família and other federal transfers via Drex;
- Programmability: composition of conditional transactions for credit and commercial contracts.
Campos Neto closed the official communication of 2023 with the phrase that would become a slogan: "Drex is the beginning of banking system tokenization". It was not just a product. It was a zero-point marker.
The GitHub controversy that no one forgot
In 2023, still in the pilot phase, the Central Bank published Drex's source code in an open repository. The Brazilian developer community detected — and disclosed — administrative functions embedded in the base contract capable of:
freeze()— freezing the balance of any address by administrator order;burn()andmint()— destroying and issuing tokens at specific addresses;- Administrative alteration of individual balance without a conventional transaction trail.
The Central Bank responded by publishing a technical note framing these functions as a reflection of "judicial powers existing in the current system" — an institutionally correct argument (bank balances today can also be blocked by court order), but which did not dispel concerns. The critical point was architectural: these functions were embedded in the base protocol, not in a separate compliance layer. What normally requires a court order, in Drex, requires only an administrative key.
The GitHub controversy became the first crack in the "innovative CBDC" narrative. It was not decentralized innovation — it was traditional infrastructure dressed in blockchain clothing.
2024: the TCU was warning
In 2024, the Federal Court of Accounts concluded an operational audit of the Drex project. The report, validated by the plenary, was courteous in form and harsh in content. The recommendations focused on four points:
- Data privacy: the chosen technology (Hyperledger Besu + Anonymous Zether) did not simultaneously meet user privacy requirements and the Central Bank's need for visibility for supervisory purposes. The architecture created an impossible trade-off;
- Performance: tests showed throughput far below what was needed to support the actual volume of Brazil's banking system;
- Scalability: the permissioned network did not scale linearly as participating nodes were added;
- Costs versus measurable benefits: the project's priority use case did not justify the investment in blockchain infrastructure given the centralized alternative.
The TCU required semi-annual progress reports from the Central Bank, instituting a monitoring regime that, in practice, created pressure for continuous re-evaluation of scope. It was not a veto. It was a demand for proof of value that the project, in its original configuration, had no way of delivering.
When the Chainalysis framework for blockchain selection in institutional tokenization was published in May 2026, it became retrospectively clear: the Central Bank had bet on the wrong archetype for the intended use case. The framework's five questions — institutional liquidity, illicit exposure, compliance tools, atomic settlement and regulatory dependency — pointed toward public infrastructure (Ethereum or equivalent L2) for the original thesis of inclusion and tokenization. The choice of permissioned DLT minimized sovereign risk, but eliminated the network gain from open tokenization. The TCU, in more sober language, said the same thing.
2025: Galípolo recalibrates the vocabulary in two stages
Gabriel Galípolo took over as Central Bank president in January 2025 and, within a few months, made it clear that Campos Neto's Drex would not survive intact. The recalibration was done in two linguistic stages, both at public forums.
First stage — Blockchain Rio, early 2025. Galípolo, on a panel, said the phrase that stopped the ecosystem: "The first version of Drex will not use blockchain". Technically, the statement did not cancel the project. Practically, it stripped it of its identity. A CBDC without blockchain is, by BIS literature's own definition, a centralized digital account — which Brazil has had since 1985, in the SPB.
Second stage — February 2025, in an interview. Galípolo took the second step: "Drex is not a digital currency, it is a tokenization infrastructure". The formulation openly contradicted the classification as a CBDC that the Central Bank itself maintains on the project's official page. But it established the new institutional framing: Drex is not money, it is infrastructure.
In August 2025, the Central Bank formalized the reformulation. Drex would focus on "lien reconciliation as the first delivery", with launch scheduled for the second half of 2026. Blockchain and tokenization were pushed to "future phases" — institutional euphemism for schedule indefiniteness.
Drex 2026: what remains in practice
The delivery scheduled for the second half of 2026 has a very specific functional perimeter:
- Who uses it: financial institutions authorized by the Central Bank. Notaries and credit platforms may be integrated in later phases — not in the MVP;
- What it does: automated lien consultation on assets used as collateral. The system responds with a binary "has/does not have lien", releasing or blocking the registration of new collateral;
- How it works: centralized API hosted on Central Bank infrastructure, integrated with SPB and Pix, with governance and audit layer under exclusive regulator control;
- What it doesn't do: it does not issue tokens, does not operate smart contracts, does not run blockchain, does not have a citizen wallet, does not replace physical real or digital real in a checking account.
The priority use case — unlocking capital stuck in collateral duplication in credit operations — is legitimate and has measurable value. Febraban research estimates the annual cost of credit operations interrupted by lien uncertainty at tens of billions of reais. Reconciling this layer produces real savings. But it requires no blockchain, requires no tokenization, and requires no rebranding as "CBDC". It was a traditional interbank system all along.
The gap that the private sector filled
While Drex was transforming into a back-office tool, the gap in real tokenization was filled by the regulated private sector. Three simultaneous fronts:
The B3 front. The real-backed stablecoin from the Brazilian exchange, announced in May 2026 along with Central Bank Resolution 519, offers what Drex promised and did not deliver: a digital real with institutional purpose, backed 1:1, operated on regulated infrastructure, integrable with other tokenized assets.
The USD stablecoin on Pix front. The integration of USDC to Pix by selected institutions, which we covered in April, created, in practice, a shadow Drex in tokenized dollar. The Brazilian citizen can move stablecoin via Pix in real time — exactly the use case that the original Drex promised for digital real.
The agribusiness RWA front. The Brazilian chapter of RWA that Chainalysis identified in the 2026 report shows that agro tokenization — soy, cattle, crop receivables — is being built on public infrastructure (Polygon, Avalanche) by private players like Liqi, B3 and trading companies. Drex is not part of this equation.
The global RWA map of the first quarter of 2026 does not mention Brazil among jurisdictions with relevant volume of tokenized treasury, digital gold or on-chain commodities. Institutional leadership, in the BlackRock BSTBL/BRSRV trilogy, stablecoins in stock of US$ 321 billion and American treasury tokenization, lies in the field of regulated privates. Brazil was building public CBDC when the world was building private tokenized treasury.
The geopolitical context that surrounds the end
While Drex shrinks to a lien tool, Europe hardens its anti-dollar-stablecoin discourse — with Lagarde's statement in Madrid explicitly endorsing the Central Bank's blockade against dollar stablecoins — and the Bank of France accelerates the digital euro. The US, through the CLARITY Act and GENIUS Act, chose to delegate dollar tokenization to the regulated private sector (Circle, Tether, BlackRock BUIDL).
Three routes for central bank digital currency, three distinct institutional choices:
- EU: public CBDC (digital euro) + blockade on private stablecoin;
- US: rejection of public CBDC + endorsement of private stablecoin via GENIUS/CLARITY;
- Brazil: neither one nor the other. CBDC emptied for banking back-office + regulatory blockade on external stablecoin via Resolution 561. The gap in real tokenization is left to private initiatives (B3, fintechs) without their own framework.
The point is not that Brazil's route is wrong. It is that it is not a route — it is a retreat. Drex ceased to be a flagship project without the corresponding state alternative being formalized. The regulated private sector fills the gap in fact, but without explicit tokenization framework that gives it scale. It is reactive, not designed.
The ON3X perspective
Three readings for the final act of Drex and what it means for the Brazilian ecosystem in the next 18 months.
1. Drex 2026 delivers real value, but it is not what was announced in 2023. Lien reconciliation is a legitimate and measurable use case. But calling this delivery "Drex" keeps alive an institutional narrative that no longer matches the scope. The Central Bank would gain in transparency by renaming the product — something like "NGL: National Lien Registry" — and formally archiving Drex as a CBDC for future resumption. Brand inertia produces expectations that delivery cannot meet.
2. The TCU was more technical, earlier, than the Central Bank's official communication. The 2024 audit already pointed out, in sober language, what three years later became silent consensus: the chosen technology did not serve the intended use case. That it took almost two years for the Central Bank's Board to translate this conclusion into a scope revision is the kind of institutional lag that costs dearly in the currency of innovation. The semi-annual reports required by the TCU become, retrospectively, the most relevant instrument of project governance.
3. The gap in real tokenization was left open for the regulated private sector — and this can be good or bad, but it needs framework. B3's stablecoin is an intelligent institutional response to the gap. USDC-Pix integration is an ingenious operational response. RWA tokenization in agribusiness is a market response. But the sum of the three does not substitute for an explicit public policy on real tokenization. Without framework, Brazil operates in a limbo: the emptied CBDC does not tokenize, the private real-backed stablecoin still does not have clear regulation, and the citizen ends up preferring tokenized dollar to tokenized real. Digital monetary sovereignty, which the Central Bank publicly defends as motivation for Resolution 561, depends exactly on the framework that the disappearance of Drex left open.
The vanguard announced in 2023 was meant to be Brazilian. The 2026 delivery is discrete, technical and necessary — but loses the adjective. Drex, as a popular CBDC, died silently between Galípolo's announcement at Blockchain Rio and the first page of the operational plan for the second half of 2026. What remains now to live ahead: private tokenized real, with or without explicit regulator framework.
Frequently Asked Questions
What is Drex in 2026?
Drex 2026 is a centralized lien reconciliation infrastructure operated by the Central Bank. It allows financial institutions to consult, via API, whether an asset already has a lien registered elsewhere before accepting it as collateral in a credit operation. It does not operate on blockchain, does not issue tokens, has no public access and does not replace physical real or real in a checking account.
Does Drex use blockchain?
No. The original version of the project, announced in 2023, planned to use Hyperledger Besu (permissioned DLT based on Ethereum). In 2025, the Central Bank president, Gabriel Galípolo, publicly stated that "the first version of Drex will not use blockchain". The 2026 delivery is centralized, without DLT.
Was Drex canceled?
Not formally. The Central Bank maintains the name "Drex" for the 2026 delivery, but the scope was reduced from a retail CBDC with native tokenization to a lien reconciliation infrastructure restricted to financial institutions. Blockchain and tokenization were left for "future phases" with no public timeline.
Who can use Drex in 2026?
Only financial institutions authorized by the Central Bank. Direct public access (from citizens, via wallet) was discarded. Notaries and credit platforms may be integrated in later phases, but are not in the MVP scope of 2026.
What did the TCU say about Drex?
The Federal Court of Accounts concluded an operational audit of the project in 2024 and pointed out flaws in privacy, performance, scalability and cost-benefit of the chosen blockchain architecture. The TCU recommended adjustments, required semi-annual progress reports from the Central Bank and established continuous monitoring. The reformulation announced by the Central Bank in 2025 and 2026 follows these recommendations.
What about real tokenization without Drex?
The gap in real tokenization is being filled by the regulated private sector: the real-backed stablecoin announced by B3 in May 2026, the integration of USDC to Pix by selected institutions, and tokenization initiatives for agribusiness assets (Liqi, trading companies, B3). What is lacking is an explicit regulatory framework for real tokenization — something Drex promised to deliver and will not.