CBDCs: what are central bank digital currencies

The money you use every day is already almost entirely digital. Your salary lands in your account; purchases go out by card or transfer. Even so, there is an important difference between this money and a banknote. The banknote is issued by the central bank. The balance in your account is a promise from your bank.

A central bank digital currency, or CBDC, would be a third form: digital money issued directly by the central bank. This article explains the concept, the models under discussion, the risks and the stage of some projects as of October 2026.

What a CBDC is

The Bank for International Settlements (BIS), the organization that brings together central banks, defines a CBDC as digital money denominated in the national unit of account that is a direct liability of the central bank (BIS, Annual Economic Report, June 2021). The January 2025 U.S. executive order, which approaches the subject from another angle, uses practically the same definition (The White House, January 2025).

Three terms help explain why this matters:

• National unit of account. A CBDC is worth exactly one unit of the country’s currency. A digital euro would be a euro, not a token whose price fluctuates.
• Direct liability. Whoever holds the CBDC has a claim on the central bank, not on a commercial bank. The same is true of a banknote.
• Central bank money. This is the money the central bank itself issues. Today, the public only has access to it on paper. Banks have digital access, through the reserves they hold at the central bank.

That is why a CBDC is not the same as instant payments. Systems such as Pix (Brazil’s instant payment system) or India’s UPI move bank deposits quickly, but they do not create a new form of money (see the main instant payment models around the world).

Nor is it the same as a stablecoin. Stablecoins are issued by private companies and depend on the reserves they hold, as explained in what are stablecoins. A CBDC is issued by the state and does not depend on third-party reserves.

Retail and wholesale

A 2018 report by the Committee on Payments and Market Infrastructures (CPMI), linked to the BIS, split CBDCs into two types (CPMI and Markets Committee, March 2018):

• Retail CBDC (or general purpose): available to the public. This is the one that usually appears in the news, described by the BIS as something close to a digital form of cash.
• Wholesale CBDC: restricted to financial institutions, to settle transactions among them. It works much like today’s bank reserves, but can enable conditional payments, which only complete when another condition is met (for example, delivery of a security).

The distinction is important because the debates are different. A wholesale CBDC affects the infrastructure of the financial market. A retail one affects the relationship among the public, banks and the central bank.

How a retail CBDC would be designed

The BIS discussed the design choices in its 2021 annual report (BIS, June 2021). Two of them appear in almost every project.

Who serves the public. The central bank could serve each user directly, but the BIS considered this model undesirable: it would concentrate costs and tasks at the central bank and reduce private innovation. The most promising direction, according to the report, is the two-tier model: the central bank issues the currency and maintains the core infrastructure, while banks and other institutions handle opening wallets, serving customers and carrying out anti-money laundering checks.

Account or token. In an account-based design, access is tied to the user’s identity, as in a bank account. In a token-based design, value circulates more like a banknote, validated by cryptographic credentials. The report assesses that full anonymity would hardly serve the public interest, because of the fight against financial crime.

Privacy. Since a CBDC records payments, the subject is central. The BIS advocates separating who provides the payment service from who controls the data, and limiting authorities’ access to exceptional cases provided for by law.

The most discussed risk: deposit flight

If people can keep central bank money on their phones, some deposits may leave the banks. The 2018 CPMI report already warned that, in times of stress, deposits could flee more easily to the central bank (CPMI, 2018).

The risk has two sides:

• Day to day, fewer deposits can mean less money for banks to lend, or more expensive credit.
• In a crisis, a digital run could be faster than a run on bank branches.

The responses most cited by the BIS are limits on how much each person can hold, low or zero remuneration, and a mechanism that automatically sends anything above the limit to the user’s bank account (BIS, 2021).

Other risks come up in public debate: surveillance of spending, cyberattacks on critical infrastructure and the cost of building something that competes with private payment methods that already work well. Each country weighs these risks differently, as the cases below show.

How many countries are studying, how many have launched

There are two reference sources, with different methods.

The BIS survey carried out in 2024 and published in August 2025 covered 93 central banks. Of these, 91% (85) were exploring a retail CBDC, a wholesale CBDC or both. Wholesale projects were generally more advanced than retail ones. More than a third of jurisdictions said they had accelerated their work because of stablecoins and other crypto-assets (BIS Papers No. 159, August 2025).

The tracker of the Atlantic Council, a U.S. think tank, monitors projects independently. In its May 2026 update, it counted 146 countries and currency unions exploring a CBDC, together accounting for more than 98% of world GDP. Only three had launched a retail CBDC: the Bahamas, Jamaica and Nigeria. Another 41 projects were in the pilot stage (Atlantic Council, CBDC Tracker). These are data from a private research organization, not from an official authority.

The gap between “studying” and “launching” is the most revealing figure. Almost every central bank follows the subject. Very few have concluded that it is worth putting a retail CBDC in the hands of the public.

The most closely watched cases

Euro area: digital euro. The European Central Bank (ECB) has not yet decided to issue the digital euro, but moved to the next phase of the project in October 2025. The ECB’s timeline foresees a pilot starting in the second half of 2027 and readiness for a possible first issuance in 2029, if European legislation is approved by the end of 2026 (ECB, digital euro). According to the ECB’s FAQ, the EU Council set its position in December 2025 and the European Parliament in July 2026, and negotiations between the institutions were under way. The proposed design includes legal tender status, distribution through banks and other providers, free basic use, offline payments with privacy close to that of cash, and no interest on balances. The ECB tested hypothetical holding limits of up to €3,000 per person, but the final amount had not been set (ECB, FAQ).

China: e-CNY. The Chinese pilot is the largest in the world, according to the Atlantic Council. By the end of November 2025, the e-CNY had recorded a cumulative 3.48 billion transactions worth 16.7 trillion yuan. As of January 1, 2026, the People’s Bank of China changed its framework: e-CNY held in commercial bank wallets came to be treated as a bank deposit, earning interest under deposit rules and covered by deposit insurance (State Council of China, December 2025). In practice, e-CNY held at banks stopped being a direct liability of the central bank, which moves it away from the classic definition of a retail CBDC.

United States: prohibition. An executive order signed on January 23, 2025 prohibited federal agencies, except as required by law, from taking any action to establish, issue or promote a CBDC in the U.S. or abroad (The White House, January 2025). According to the Atlantic Council, the U.S. is the only G20 country not exploring a CBDC.

United Kingdom: decision pending. The Bank of England and HM Treasury said they would announce in 2026 whether to proceed with a digital pound. In the most recent update of the bank’s news page, from April 2026, the decision had not yet been announced (Bank of England).

Brazil: Drex. The Atlantic Council tracker treats Drex, a project of the Banco Central do Brasil (Brazil’s central bank), as a tokenized credit pilot within the movement toward wholesale infrastructure, not as a digital version of the cash real for the public. According to a November 2025 report by Finsiders Brasil, the Central Bank shut down Drex’s blockchain-based test platform after concluding it did not meet privacy requirements, and began working on a new infrastructure (Finsiders Brasil, November 2025). This information comes from specialized press; the editorial team did not find an official statement with these details.

What this changes for you

For now, almost nothing in Brazil. There is no retail digital real in circulation. In the editorial team’s assessment, Pix already delivers much of what a retail CBDC would promise in speed and cost, which reduces the pressure for one.

If you live or do business in countries with advanced projects, three questions help evaluate any proposal: who will see your payment data, how much you will be able to hold and whether there will be interest. Also be wary of offers of a “government digital currency” outside official channels. Scams often use the names of real projects.

Key takeaways

• A CBDC is digital money issued directly by the central bank, in the national currency. It is not a crypto-asset, not a stablecoin and not an instant payment system.
• Almost every central bank is studying the subject, but only three countries had launched a retail CBDC as of May 2026, according to the Atlantic Council.
• The debate revolves around privacy, bank stability and real usefulness. Choices vary widely: the euro area is moving forward, the U.S. has banned it, China changed its model and Brazil reshaped Drex.