Colombia Court Demands Algorithm Transparency, Implementation Unclear
Colombia’s Constitutional Court demands algorithmic transparency, ruling citizens have a right to access government algorithm source code.

Colombia’s Constitutional Court has taken a firm stance on “algorithmic opacity”, linking it to human dignity and labeling it a threat to democracy. The court’s T-067 ruling granted a citizen access to the source code of CoronApp, a government pandemic app, setting a precedent that public information includes state algorithm operations.
The court defined algorithmic transparency as “information availability about algorithm systems to understand their operation and evaluate performance.” This includes accessibility (allowing external audits) and explainability (presenting results in simple language). The Joint Directive 007 adopted this definition, though it misattributed the concept’s origin to the court.
The Challenge of Implementation
The court’s definition remains conceptual, lacking specifics on information quantity, audit standards, or language simplicity. This gap persists in other rulings, like T-323, which mandated AI regulation in the judiciary without setting standards. The T-008 ruling reiterated the need for checks on AI in public functions, but the practical application of these concepts remains unclear.
Three Causes of Algorithmic Opacity Identified
Researcher Jenna Burrell identifies three opacity causes: deliberate secrecy, technical illiteracy, and inherent model complexity. Colombian jurisprudence treats these as a single issue, solvable with a general transparency standard. However, the core problems—proper sourcing, decision motivation, and due process—are long-standing legal challenges amplified by AI, not created by it.
Who Will Fill the Operational Void?
The Congress, with direct democratic legitimacy, could legislate, but any law would face court scrutiny for constitutionality. Judges could set standards case-by-case, but the judiciary has shown slow adaptation to technological changes, as seen during the pandemic. The Executive branch has acted faster, issuing Conpes 4144 and Joint Directive 007, but with less public debate.
Alternatively, private developers, technically best equipped to explain systems, may fill the void, though they have less incentive to do so. This leaves legal professionals in a bind: who to trust more—the state or private developers? Colombia has yet to decide, and this indecision defines the practical reality of algorithmic opacity in its legal system.
Regulators Need Unique, Case-Specific Approaches
The issue isn’t the lack of a definition but the absence of regulators who understand that each case requires a unique approach. The law may fall short without technologically literate officials and interdisciplinary lawyers who can bridge the gap between concept and practice.
The court’s rulings have set a theoretical framework, but the operational details remain unresolved. As AI tools like ChatGPT become more prevalent in legal work, the need for clear, practical standards grows. The focus now shifts to transparency and accountability, with questions about who will define and enforce these principles in a rapidly changing technological environment.
For lawyers using AI, uncertainty surrounds liability and accountability. Without clear standards, they handle a system where rules are still being written, often case-by-case. This ad hoc approach may not keep pace with technological advancements, leaving legal professionals and the public in regulatory limbo.
The court’s emphasis on human dignity and democratic principles is clear, but achieving these goals in practice remains uncertain. As Colombia addresses these questions, the practical implications for legal professionals and the public are significant. The interdisciplinary nature of the challenge requires both legal expertise and technical understanding, a shift the legal profession is still adapting to.
Practical Challenges in Achieving Transparency
Meanwhile, the black box nature of algorithms poses challenges. While the court has acknowledged the problem, the solution is still evolving. Next steps likely involve legislative action, judicial interpretation, and public engagement to ensure algorithmic transparency becomes a practical reality.
The Joint Directive 007 and court rulings have laid groundwork, but operational details are missing. Moving forward, the focus is on implementation and enforcement, ensuring transparency and accountability principles are consistently applied across government and private sectors.
The legal profession plays a key role, requiring lawyers to understand both law and the technology behind algorithms. This shift to interdisciplinary practice is underway but demands ongoing education to keep up with technological advancements.
The goal is to ensure algorithmic systems are transparent, accountable, and fair. While the court has taken an important step, the journey is just beginning. Challenges are significant, but so are the opportunities to create a more just and equitable system.
The T-067 ruling and directives set the stage, but the real work lies ahead. As Colombia handles this complex field, its lessons will be valuable globally as others grapple with algorithmic decision-making in legal systems and beyond.
Collaborative Effort Needed for Just System
Addressing algorithmic opacity requires a collaborative effort involving all stakeholders, judiciary, legislature, private sector, and public. Only through such collaboration can Colombia achieve a system that is truly transparent, accountable, and just.
Jenna Burrell’s research highlights the complex nature of algorithmic opacity, from secrecy to technical complexity. Each cause requires a distinct solution, yet Colombian law treats them as one issue. This oversimplification risks regulatory gaps, especially as AI advances rapidly.
The Congress, though democratically legitimate, has been slow to act, leaving the Executive branch to lead with initiatives like Conpes 4144. However, these lack the public debate and scrutiny of legislative action, raising questions about their legitimacy and effectiveness.
The judiciary faces challenges in setting standards case-by-case, a time-consuming process that may not keep pace with technology. The lack of technical expertise among judges and lawyers compounds the issue, making informed and fair decisions difficult.
The private sector, often algorithm developers, has conflicting interests in transparency. While they have the expertise to explain systems, competitive reasons may hinder this. Balancing transparency with market competition is key to serving the public interest.


