When Law Becomes Data: What Brazil’s LexML Reveals About Akoma Ntoso
A Conversation with Monica Palmirani on Building and Sustaining Digital Systems
BY BEATRIZ REY
The first time I used Brazil’s LexML portal, I was trying to reconstruct the life of a bill. Who had introduced it. When it had changed. How it had moved between committees and the floor. LexML helped. And it did not.
Within seconds, I could access the official text, related legislation, and links to publications from the Presidency, Congress, and the National Press. Everything was neatly organized and searchable.
But one thing was missing. There was no easy way to see who had sponsored the bill. Which party had pushed it forward. The political story of the law, the part that explains why it exists, was largely absent. Even though Brazil adopted Akoma Ntoso, the open standard that powers LexML, the technical standard can only bridge institutional silos if the implementing parliaments prioritize that in its deployment.
What is Akoma Ntoso?
Akoma Ntoso was conceived as a technical standard designed to bring coherence to fragmented legislative and judicial information systems. As Monica Palmirani, one of the architects of the standard, explained to me, it is “a legal XML vocabulary for modeling the legal knowledge and documents of the legal domain from [a] parliamentary act [or] judiciary act.”
It was originally developed in 2008 with United Nations funds for Africa to enable the digitalization and interconnection of parliamentary documents across different legal traditions, languages, and institutional structures. According to Palmirani, the standard was developed to serve “all the legislative systems from common law to civil law,” including unicameral and bicameral legislatures and multilingual contexts such as South Africa, the European Parliament, and the United Nations.
In this sense, Akoma Ntoso was not designed as a political analytics tool, but as an infrastructure for preserving legal documents in a stable, interoperable, and future-proof format. Palmirani emphasizes that its openness and reliance on XML allow it to remain readable across technological changes, preventing legal records from becoming inaccessible as software evolves.
Beyond standardizing documents, Akoma Ntoso is built to capture the procedural life of legislation. According to Palmirani, the standard incorporates “many metadata for tracking the versioning of the legislation over the time” and “for tracking the workflow inside of a procedure of the Parliament.” This metadata makes it possible to reconstruct how a bill moves through committees, amendments, debates, and votes.
One of its central objectives, she explains, is “to track the history of the law from the bill to the assembly, to the committees, and to the amendments and to [the] vote.” By preserving this institutional memory, the standard allows citizens, judges, and researchers to understand “what happened to the document” and why a final law differs from its original proposal. Akoma Ntoso thus operates as a structured archive of legislative processes, enabling transparency and accountability by making procedural histories visible — while remaining limited to what institutions themselves are willing and able to document.
Akoma Ntoso has spread quietly and unevenly over time. European parliaments, such as Italy’s Chamber of Deputies and Senate, adopted it for archival purposes. Constitutional courts used it to structure judgments (one example is also Italy’s High Court of Cassation). International organizations like mySociety experimented with it for normative databases. Brazil built LexML on top of it.
Yet, as Palmirani notes, there was never a single center coordinating these efforts. Every day, she says, she discovers that other people are using it, often without systematic coordination. Knowledge circulates through informal networks, conferences, and bilateral exchanges rather than through centralized governance structures. This decentralized diffusion reflects both the openness of the standard and the fragmentation of legislative IT ecosystems.
The Organizational Challenge
One of the most striking aspects of Palmirani’s account is how little she talks about software. For her, the central challenge of Akoma Ntoso has never been technological. It is a transformation of the organization, she insists. Implementing the standard requires rethinking how laws are drafted, reviewed, archived, and published. It demands coordination between lawyers, clerks, IT staff, and administrators. It also requires long term institutional commitment. And it takes time.
In the first three years, Palmirani notes, you probably do not see many benefits. That delay is politically dangerous. Legislators and managers often expect immediate results. When they do not materialize, support erodes. Projects stall. Teams are reassigned. Budgets disappear. Many digital reforms fail not because the technology is flawed, but because the institutions are unstable.
Going back to Brazil: In 2010, Brazil became one of the first countries in the world to adopt Akoma Ntoso. Even more unusually, it adapted the standard to Portuguese, becoming the only country to implement it in a language other than English. The result was LexML, a national platform designed to bring together legislation, court decisions, and bills in a single searchable system. Its interface is deliberately simple: a search bar, a few filters, and results that can be refined by institution, date, document type, or jurisdiction.
Seen from the outside, LexML looks like a success story. And in many ways, it is. It centralizes access, connects institutions, and preserves official records, reducing informational asymmetries. For researchers like me, it has become indispensable. But its limits are just as revealing. LexML is less effective at organizing political information. One key reason is institutional. In Brazil, the databases of the Chamber of Deputies and the Federal Senate are not integrated. Each house maintains its own systems, metadata standards, and workflows. This fragmentation has never been fully resolved.
As a result, LexML reflects these organizational boundaries. While it successfully aggregates official texts from multiple institutions, it does not integrate core political data such as bill sponsorship, party affiliation, or bill status. Those data remain siloed in separate parliamentary databases. This is not a technical failure of Akoma Ntoso. It is an institutional one: digital standards cannot integrate what political organizations themselves keep separate.
Akoma Ntoso in the Age of Artificial Intelligence
When I asked Palmirani about artificial intelligence, her answer did not begin with algorithms or automation. It began with infrastructure. Over the past few years, Akoma Ntoso has been quietly rediscovered in the context of AI. Suddenly, the structured legal data that once seemed like a niche concern has become central to new systems for automated drafting, consistency checks, semantic search, and regulatory monitoring.
From this perspective, the past is catching up with the present. What was built two decades ago for transparency and interoperability is now becoming the foundation for machine assisted governance. Palmartini sees enormous potential in this shift. Akoma Ntoso, she argues, can support drafting, help detect errors, and monitor the evolution of the law at scale. Properly used, it can make legislative work more coherent and more reliable.
But she is also wary. “There is the trend to skip any annotation nowadays and to use LLM generative AI from the plain text. This is a very shortcut approach, but full of risk because we lose the legal tradition, the legal knowledge, the legal accumulation for a century of the knowledge of the legal tradition of the country,” she says.
For Palmirani, meaningful digital reform requires institutional patience and sustained commitment. “It’s important to have a strong commitment and continuity, because in the first three years probably you don’t see some benefits… but after all, you can see that the database is ready for communicating with all the rest.”
Brazil’s LexML shows how difficult that task is. Digital standards do not simply modernize legislatures – they mirror existing institutional arrangements. In this sense, Akoma Ntoso’s, and digitalization processes more broadly, greatest challenge may be ensuring that this grammar is embedded in institutions strong enough to sustain it.
Modern Parliament (“ModParl”) is a newsletter from POPVOX Foundation that provides insights into the evolution of legislative institutions worldwide. Learn more and subscribe at modparl.substack.com.
