Blockchain against record fraud: the real limits

Record fraud is an old problem: someone alters a property title, forges a signature or deletes an entry to take over an asset or hide an expense. The promise of blockchain seems tailor-made for this: a record that no one can change in secret.

The question is whether that promise holds up in practice. This article separates what the technology actually guarantees from what it cannot reach, looks at real projects in Georgia, Sweden and Brazil, and ends with criteria for evaluating proposals. If you would first like to understand how a blockchain chains blocks together, see how the Bitcoin blockchain works.

What blockchain guarantees

The technology’s strong point is integrity: once a record is in a published block, changing it without leaving a trace becomes very difficult. NIST, the U.S. standards institute, prefers more precise terms than “immutable”: blockchains are tamper evident (a change shows) and tamper resistant (changing is costly or difficult) (NIST IR 8202, 2018).

The same report explains why “immutable” is an overstatement:

• On open networks, the most recent blocks can be replaced by an alternative chain, and an attacker with enough resources can rewrite part of the history.
• On permissioned networks, those in which an owner or consortium decides who participates, administrators may, depending on the configuration, rewrite the block history.

In other words, the protection depends on who controls the network. This matters a great deal for public records, which almost always use permissioned networks.

What it does not guarantee: the input data

The central limitation has a name in computing: garbage in, garbage out. If false data is recorded, the blockchain preserves the false data just as efficiently as it would preserve true data.

NIST describes this as part of the oracle problem: a blockchain can record data entered by people or sensors, but there may be no way to know whether it reflects what happened in the real world. People can record false information, on purpose or by mistake. Ethereum’s documentation gives the same warning about oracles, services that bring external data onto the network: you have to trust that the data came from the right source and was not tampered with along the way (ethereum.org, oracles).

The warning comes even from advocates of the technology. In a 2017 issue of the international registrars’ journal IPRA-CINDER, executive Matt Regan, who favors using blockchain in property registries, cited “garbage in, garbage out” and acknowledged that blockchain “does not solve” the challenge of producing correct documents; that remains the job of whoever verifies the parties’ identities and drafts the instruments (Regan, IPRA-CINDER, 2017).

In practice, this splits fraud into two groups:

• Fraud after recording. Someone alters or deletes an entry that has already been made. Here blockchain genuinely helps, because the change becomes evident.
• Fraud at entry. A deed signed with a forged power of attorney, a sham invoice, a registration made with a stolen ID. Here blockchain does not help: it records the fraud and makes it hard to correct.

A property registry is not just a ledger

Economist Benito Arruñada, of Pompeu Fabra University, published a critique in the same journal titled “don’t expect too much”. His main argument is that calling a property registry a ledger is misleading (Arruñada, IPRA-CINDER, 2017).

According to him, registries of rights, such as those of Germany, Spain and Australia, do not merely timestamp and store documents: before recording, they check whether the transaction respects the rights of all other holders of rights in that property. That legal analysis, Arruñada writes, is not easy to automate. For that reason, he argues that the technology is more likely to advance in document authentication and data archiving than in replacing registries. For a blockchain entry to be effective against everyone, a public authority would still be needed to define its value as evidence.

This is an academic position, not a consensus, but it helps separate two things that are often mixed up: storing a document securely and deciding who the owner is.

The case of Georgia

Georgia is the most frequently cited example. According to a press release from Bitfury, the company that built the system with the National Agency of Public Registry (NAPR), the two signed an agreement in April 2016 for a pilot project to record land titles (Bitfury, press release).

The design, as described in the release, is this: a private, permissioned blockchain integrated with the agency’s digital system and anchored to the Bitcoin blockchain through a timestamping service. Citizens receive a hash of the document, a kind of mathematical fingerprint, which is published on the Bitcoin network and can be checked when needed. A 2020 summary by the think tank New America describes the system as an internal complement to the existing registry, with the user-facing part left unchanged, and lists services such as sales, mortgages, leases and notarial services (New America, 2020).

Two details matter when assessing the case:

• The registry already worked. The release itself highlights that Georgia ranked among the top three countries for ease of registering property in the World Bank’s Doing Business ranking. Blockchain was added to a trusted digital registry, not used to create one.
• The agency still decides. The NAPR is the one that verifies and signs the document. The blockchain proves that the document existed in that form on that date; it does not prove that the transaction behind it was legitimate.

Our newsroom did not find a recent, independent evaluation measuring whether the system reduced fraud in Georgia. The available claims about results come largely from the company and from entities linked to the project.

The case of Sweden

In Sweden, the land registration and cadastral agency, Lantmäteriet, ran a proof of concept in 2016 and a test environment that ended in March 2017, according to Computer Weekly. The solution used a private blockchain and smart contracts developed by the company ChromaWay, with the participation of Telia, the consultancy Kairos Future and the banks SBAB and Landshypotek (Computer Weekly, July 5, 2017).

The agency’s head of development, Mats Snäll, described the test as small-scale and said it would still need to be tested with more partners and more transactions. He also pointed to network governance and regulatory issues to be resolved before large-scale use. The projections of cutting the time between the purchase contract and registration from four months to a few days, and of saving more than 100 million euros a year, were the agency’s estimates, not measured results.

In 2018, Kairos Future announced a live test in which a property purchase was carried out entirely digitally during a seminar, now with the participation of the Swedish Tax Agency (Skatteverket) and a real estate agency (Kairos Future). The savings figures cited in that announcement are projections by one of the partners.

Arruñada notes that the Swedish proposal mainly reorganized the contractual stage that precedes registration, with the registry keeping, at least initially, the power to review and decide. Our newsroom did not find an official source confirming that the system was adopted in production.

And in Brazil

In Brazil, public registries exist, under Law 6.015/1973, to give “authenticity, security and effectiveness” to legal acts (Law 6.015). The modernization under way is electronic, but it does not depend on blockchain:

• ONR. Law 13.465/2017 required the Electronic Real Estate Registry System (SREI) to be implemented and operated nationwide by the National Operator of the Electronic Real Estate Registry System (ONR), a private nonprofit entity, with the National Justice Inspectorate of the CNJ (Brazil’s National Council of Justice) as regulator (Law 13.465, art. 76).
• Serp. Law 14.382/2022 created the Electronic System of Public Registries, whose goals include electronic recording of acts, interconnection of registry offices, online service and searches for asset freezes and liens. Regulation is the responsibility of the National Justice Inspectorate (Law 14.382). The text of the law does not mention blockchain.

Notary offices, which draw up deeds and authenticate documents, have used their own blockchain since 2020. According to the Colégio Notarial do Brasil (the national association of notaries), the Notarchain network, linked to the e-Notariado platform, is permissioned, based on Hyperledger Fabric, and every node must be a notary office. The association reports that the share of digital notarial acts rose from 1% in 2020 to 42% in 2024 (CNB/CF, 2024). The figures come from the very entity that operates the system.

Notaries themselves point out the technology’s limit. In a Valor Econômico report reproduced by the São Paulo notarial association, the section’s vice president, Andrey Guimarães, asked: “The information is immutable, but without prior attestation, is it true?” (CNB/SP, 2023). It is worth remembering that the profession has a direct interest in the topic.

For real estate, there is an additional issue: a token does not transfer ownership, which in Brazil depends on registration in the property’s record (matrícula). The inspectorates’ decisions on the matter are covered in does a token represent legal ownership of a property?.

What about public spending?

The same logic applies to the idea of putting government invoices, payments and contracts on a blockchain. In the Token & Tech segment that accompanies this article, that is exactly the discussion: whether immutable records, transparency and traceability of a city government’s spending would reduce corruption and increase public trust.

In our newsroom’s assessment, based on the sources above, the technology can help prevent a payment from being deleted or altered afterward and make it easier for anyone to audit. It does not, however, prevent an overpriced invoice, a project that was never delivered or a bid rigged before anything was recorded. Those frauds happen off the network and still depend on oversight.

How to evaluate a “blockchain record” proposal

Before accepting that a project reduces fraud, it is worth asking:

1. Which fraud does it target? Later alteration of records or false data at entry? Only the first is blockchain’s strong point.
2. Who checks before recording? If no one verifies identity, documents and third-party rights, the fraud is just better stored.
3. Who controls the network? On a permissioned network, administrators concentrate power, including the power to rewrite history, according to NIST.
4. Does the record have legal value? In Brazil, what counts for real estate is the matrícula at the competent registry, not an entry on a private network.
5. How is an error corrected? A record that is hard to alter is also hard to fix when there is fraud or a mistake.
6. Is there an independent evaluation? Be wary of results disclosed only by the vendor or the operating entity.

If the project involves programs that execute rules on their own, also understand the risks of smart contracts. And if someone offers an investment with the promise of “blockchain registration”, see how to spot crypto scams.

Key takeaways

• Blockchain makes later tampering evident and difficult. It does not make false data recorded at entry true.
• Georgia and Sweden used the technology alongside registries that already worked and that continued to decide what goes in.
• Independent evaluations of fraud reduction are scarce; many of the results come from vendors and operators.
• In Brazil, the modernization of registries follows Serp and the ONR; blockchain appears in notary offices, on a permissioned network, without dropping prior verification.
• For public spending, the same rule applies: tracking helps auditing, but it does not replace oversight of what happens off the network.