Colombia

Though the mechanism was originally intended to act as a decision-making tool, it now acts as a trained search engine to aid judges and court clerks in identifying important cases from the large volume of tutela filings, but retaining the role of the human judge as the sole decision-maker. There is, however, potential for the expansion of its capabilities. The implementation of PretorIA faced challenges such as digitising paper-based files and refining the algorithm, but is now used in tandem with the traditional case-by-case review process. 

PROMETEA, discussed above, is also used in Colombia to assist in the selection of tutela petitions by automatically reading large volumes of filings and flagging urgent cases for priority review, but the Constitutional Court retains discretion to confirm, modify, or reject the system’s suggestions. 

Legal research, analysis and drafting support 

Judges report unofficial use of commercial and general purpose AI tools, such as ChatGPT and CoPilot, for drafting and summarisation tasks, including that of public hearings. These judges have also reported achieving higher levels of productivity and being able to resolve more cases in less time.

As at March 2026, the judiciary is conducting several proofs of concept as part of its AI adoption plan, with the goal of deploying solutions that integrate AI into judicial processes and thereby improve the effectiveness of—and access to—justice services. Examples include:

• Macrojudgment analysis: aiming to improve efficiency and turnaround times in the handling of lengthy legal proceedings that currently require substantial manual work, by supporting the calculation of damages and the drafting of complex decisions, and reducing time and errors in highly complex proceedings.
• Resolution of jurisdictional conflicts: aiming to improve efficiency in the analysis and resolution of administrative jurisdictional conflicts, by streamlining processing by identifying analogous cases, and facilitating the review of records and lengthy documents.
• Semantic search and clustering in case law summaries: aiming to develop AI-enabled tools that can: (i) perform advanced searches and semantically group similar cases, and (ii) enhance jurisprudential analysis when dealing with high volumes of documentation.

Defence

As at March 2026, there are no reported cases of defence counsel in Colombia making use of AI. 

Victims 

Victims have direct participatory standing in Colombian criminal proceedings, although they are generally characterised as special participants rather than as full parties equivalent to the prosecution or defence. Under Article 11 of the Criminal Procedure Code, victims are guaranteed access to the administration of justice, the right to be heard, the opportunity to contribute evidence, information about prosecutorial decisions, and, where applicable, the right to appear before the judge of control of guarantees and to lodge remedies before the trial judge. Article 137 further provides that victims may intervene in all phases of the criminal proceeding to protect their rights to truth, justice and reparation.

AI-enabled tools are increasingly used by public institutions in Colombia to facilitate victims’ access to information, assistance services and institutional procedures.

Case management

The Bogotá Secretariat for Women has developed a chatbot designed to provide information and assistance to users seeking support in situations related to gender-based violence. The chatbot operates continuously and offers guidance on available services, including legal orientation and psychosocial assistance provided by the institution.

PERLA is a virtual assistant developed by the Government of Bogotá to respond to users’ questions and facilitate access to public services. The system operates 24/7 and allows users to obtain information about government procedures and services. Among other functions, the chatbot can connect users with the Línea Violeta, a support service for women experiencing violence.

TRAINING 

For law enforcement agencies, training is provided on an informal basis. For example, in March 2025, 25 Colombian police officers undertook a two-week training programme in India on the application of AI to citizen security. It provided greater knowledge on the strategic use of AI in public policing, using automation to detect patterns and potential risks.

There is not yet a structured training programme for judges and court staff in Colombia. However, the Superior Council of the Judiciary (Consejo Superior de la Judicatura) – the body responsible for the administration and oversight of Colombia’s judicial system – is tasked with planning the overall digital transformation of the judiciary, as well as contributing to its implementation. This includes ensuring connectivity, providing appropriate technological equipment, digitalising all judicial processes, and, finally, developing and adopting ambitious technological tools. Within this strategy, the Superior Council has been promoting the use of AI with the aim of further improving access, efficiency, and security within the administration of justice.

The Council has a Unit of Digital Transformation (Unidad de Transformación Digital). This Unit has sought to promote the adoption of a new digital culture. With that goal in mind, it has introduced several training initiatives, all available through a ‘competencies portal’, which is part of the AI strategy and aims to train judges at different levels –  from basic digital awareness to advanced technological development. The Superior Council has also organised contests and stakeholder dialogues to encourage, support, and provide feedback to bottom-up innovation proposals for applying technology in judicial work. Most of these initiatives aim to reduce judicial congestion, one of the Colombian judiciary’s pressing problems. All of this training work has been carried out in coordination and close articulation with the Rodrigo Lara Bonilla Judicial School (EJRLB), as it is another dependency of the Superior Council of the Judiciary. The EJRLB is the unit responsible for coordinating and working jointly with the UTDI on the Judicial Training Plan focused on Artificial Intelligence.

In this context, one of the key tools available through the Rodrigo Lara Bonilla Judicial School (EJRLB)—aligned with the 2025–2026 Training Plan—is its continuing education offering delivered through the School’s Academic Management System (SGA). Through this platform, training activities can be developed not only on these topics, but also on other areas of interest or specialization, including virtual courses, diploma programs, and self-paced learning modules. These learning experiences incorporate teaching strategies consistent with the School’s pedagogical model and supported by digital tools, enabling virtual delivery to contribute to the implementation of the Digital Transformation Plan across all training subprograms. In addition to initiatives within the judiciary, Colombian universities and academic institutions have begun offering specialised training programmes on artificial intelligence in the public sector and justice systems. For example, the School of Government at the Universidad de los Andes, through the Sistemas de Algoritmos Públicos project, regularly offers graduate and continuing education courses on artificial intelligence and its applications in public administration and the justice sector. These include courses such as Artificial Intelligence for the Administration of Justice: Foundations, Applications and Best Practices and Formulation of Artificial Intelligence Projects for the Justice Sector, which aim to train public officials and legal professionals on the use, governance and ethical implications of AI systems.

Training opportunities related to artificial intelligence are also increasingly available through professional and continuing education programmes. For instance, the Bogotá Chamber of Commerce (Cámara de Comercio de Bogotá) offers courses on the use of tools such as ChatGPT and other AI systems, designed to help professionals understand how these technologies can be used to automate tasks, generate content and improve digital services. Although these programmes are not specifically designed for criminal justice actors, they illustrate the growing availability of training opportunities for legal professionals and other practitioners interested in the use of artificial intelligence in professional contexts.

REGULATION 

There have been attempts to regulate the use of new technologies in the judiciary through the National Congress of Colombia; however, none of these initiatives has yet been successful. As at March 2026, there are no national laws or statutory rules that expressly regulate the use of artificial intelligence in criminal proceedings or court proceedings more generally. However, binding judicial guidelines exist (see more on the Acuerdo No. PCSJA24-12243 below), which are mandatory within the judicial branch but do not constitute statutory regulation enacted by the National Congress of Colombia. Colombia has also adopted a national AI policy and, in 2025, the government introduced a new bill (Bill No. 422 of 2025) proposing a comprehensive AI regulation. As at March 2026, this bill remains under consideration before Congress and has not been enacted (see ‘Outlook’ section below).

In the absence of a dedicated AI statute, enforcement of AI-related risks in criminal proceedings currently relies on the combined application of constitutional principles, data protection law, and general procedural safeguards. This includes, in particular, Law 1581 of 2012 and External Circular No. 002 of 2024, and the guarantees provided under the Criminal Procedure Code (Law 906 of 2004). Complementing this framework, sector‑specific institutional regulation has emerged. Notably, in 2026 the Office of the Attorney General (Fiscalía General de la Nación) issued internal guidelines on the use and scaling of artificial intelligence through Resolution No. 0077 of 2026, illustrating how AI governance in Colombia has advanced primarily through constitutional adjudication, administrative guidance, and institutional regulation rather than comprehensive legislative enactment.

Guidelines for practitioners

Judicial guidelines 

In December 2024, following a Colombian Constitutional Court ruling and a survey on the use of AI in the judiciary, the Superior Council of the Judiciary adopted Guidelines for the Responsible and Safe Use of Generative AI in the Judicial Branch (‘Guidelines’) through Acuerdo No. PCSJA24-12243. The main purpose of those guidelines is ‘to maximise the benefits and potential of [AI tools], while mitigating and managing their potential risks’. This made Colombia the first country to implement the UNESCO Guidelines for the Use of AI Systems in Courts and Tribunals. An unofficial English translation of the Guidelines is available here. 

The Guidelines of the Superior Council of the Judiciary apply across the entire Colombian judicial branch, and orient the work of around 38,000 public servants (including judges, clerks, and administrative staff) and 6,500 courts, both at the national and local level. They include categories of permitted use with exhaustive lists of permissible tasks. Using AI to perform administrative functions– such as ‘scheduling of activities or errands’, drafting minutes or administrative texts and translating documents – is always permitted. When AI is used for more substantive tasks – such as transcribing hearings, summarising the facts of a case and drafting procedural orders – it is subject to additional safeguards and requirements, including – depending on the task – ‘detailed review’ of the output and ‘special observance of transparency and accountability’. Generative AI may not be used for any task outside those listed in the Guidelines, unless expressly authorised by the Superior Council of the Judiciary. 

The Guidelines impose restrictions on both the use of certain AI tools and the type of tasks they may be used to perform: 

• Training Data Provenance: AI tools whose training data provenance cannot be verified may only be used to perform administrative support functions. 
• Evidence and legal reasoning: AI may not be used for evidence assessment, scrutiny of facts, making ‘value judgements’, and resolving ‘legal problems’.

• Automated decision-making: AI tools that apply legal rules or adopt decisions based exclusively on the tool’s responses are prohibited.

• Human rights risk: AI tools whose use could infringe fundamental rights are restricted.

• Commercial free chatbots: General-purpose or commercial chatbots in their free version may not be used for judicial purposes. 

The Guidelines set out a catalogue of 15 principles and guarantees governing the use of AI in the judiciary, including:

Enforcement of the Guidelines

The Guidelines are binding within the judicial branch and apply to judges and judicial staff. Institutional responsibility for digital transformation and technological infrastructure within the judiciary, including AI-related systems, lies with the Digital Transformation and Information Technology Unit created by Acuerdo No. PCSJA23-12130 of 29 December 2023. The Guidelines do not establish a dedicated sanctions regime for non-compliance. Oversight of AI-related risks therefore operates through existing institutional governance structures within the Superior Council of the Judiciary, as well as through constitutional review and applicable data protection enforcement mechanisms.

As of March 2026, the Guidelines continue to serve as the principal governance framework for AI in the judiciary, and the Superior Council of the Judiciary has indicated that they may evolve as technology, regulation and jurisprudence develop. To communicate the progress, challenges, and developments related to the implementation of AI in the Judicial Branch, an AI Microsite has been made available.

Criminal procedure rules 

The Criminal Procedure Code (Law 906 of 2004) and, in transitional cases, the older inquisitorial procedure under Law 600 of 2000, establish admissibility, chain of custody, and authenticity standards applicable to AI-generated or analysed evidence. While these codes predate artificial intelligence and do not refer to it expressly, their provisions on digital and expert evidence apply to outputs generated or processed by AI systems (such as facial recognition matches, predictive analytics, or PRiSMA risk assessments) when introduced in criminal proceedings. In practice, such evidence would need to comply with general standards of authenticity and reliability and may require expert testimony to support its probative value.

Additionally, provisions of the Cybercrime and Technology Misuse Laws (Law 1273 of 2009 which amends the Penal Code to criminalise unauthorised access, data interception, and damage to information systems) apply to malicious uses of AI that compromise evidence or judicial data in criminal cases.

Data protection legislation 

Law 1581 of 2012, also known as General Data Protection Law, and its regulatory decrees (Decree 1377 of 2013 and Decree 1081 of 2015), governs the protection of personal data in both the public and private sectors. Although the statute does not expressly mention AI, its provisions apply broadly to any ‘system’ that processes personal data. By extension, the requirements of Law 1581, including lawfulness, purpose limitation, security, and transparency, also govern the AI-assisted processing of personal data in judicial proceedings and criminal investigations conducted by courts and law enforcement.

The law is enforced by the Superintendence of Industry and Commerce (SIC), which can investigate and sanction improper AI-related data processing conducted by both public and private entities. The SIC issued Circular Externa 002 of 2024, clarifying privacy obligations for AI systems processing personal data, requiring privacy and risk-management safeguards for such systems, including those used in law enforcement and judicial contexts.

Human rights 

The use of AI in criminal proceedings must be consistent with procedural guarantees and fundamental rights in the Colombian Constitution, including the right to due process and equality before the law. The Constitutional Court reaffirmed in its August 2024 ruling (T-323/2024), that AI tools cannot substitute judicial reasoning and must not undermine these protections. The Court required that whenever AI is employed in adjudication, its use requires disclosure, transparency, and human accountability. 

Relevant guarantees, including the right to a fair trial and privacy rights, under other international human rights treaties to which Colombia is a party, such as articles 8 and 11 of the American Convention on Human Rights, articles 14 and 17 of the International Covenant on Civil and Political Rights or articles 16 and 40 of the Convention on the Rights of the Child, may also provide additional guidance.

Outlook

In 2019, Colombia issued its first Digital Transformation and Artificial Intelligence Policy through Conpes 3975, which aims to enhance the generation of social and economic value in the country through the strategic use of digital technologies in the public and private sectors.

Subsequently, in February 2025, the national government approved its National Artificial Intelligence Policy document (CONPES 4144). The policy is designed to promote the development, adoption, and ethical and sustainable use of AI across both the public and private sectors, drawing on international guidelines such as the OECD AI Principles and the UNESCO Recommendations on the Ethics of Artificial Intelligence.

In June 2025, Colombia advanced its AI regulatory agenda by introducing Bill No. 422 of 2025 inspired by the EU AI Act. The draft bill proposes a risk-based classification of AI systems, creation of a national oversight authority (under the Ministry of Science, Technology and Innovation), regulatory sandboxes, and binding principles such as human oversight, transparency, and privacy protection, along with sanctions for non-compliance. AI uses in law enforcement and the administration of justice are classified as ‘high-risk’. The draft also contemplates civil and criminal liability for harm from AI misuse. The Ministry of Justice and the Ministry of Science, Technology and Innovation have signalled that the justice sector will be prioritised for pilot projects under the forthcoming national AI bill. As at March 2026, the bill remains under consideration by the Colombian legislative process. 

Moreover, the Superior Council of the Judiciary is developing secondary regulations to operationalise the Guidelines for the Responsible and Safe Use of Generative AI in the Judicial Branch (Agreement PCSJA24-12243), including specific technical standards for approved AI tools.

CASES

The Colombian Constitutional Court has assessed the legality of AI-assisted drafting of judicial decisions and ruled on whether citizens have the right to access the algorithms used in public administration. In addition, there have been various decisions of Colombian courts addressing the use of AI by litigants and tribunals. 

Judicial use of AI

Early judicial practice already reflected the use of generative AI tools prior to constitutional guidance. In a January 2023 labour judgment (Sentencia No. 032, Juzgado 1º Laboral del Circuito de Cartagena), the trial judge expressly acknowledged having used ChatGPT to “extend the arguments” of a decision concerning healthcare entitlements for a minor with autism spectrum disorder. While the outcome of the case was not challenged on due process grounds, subsequent review noted that the AI tool produced imprecise legal references, including questionable citations to statutory provisions. This early example anticipates the concerns later addressed by the Constitutional Court in T‑323/2024 (see below).

In its August 2024 ruling (T-323/2024), the Constitutional Court addressed for the first time the use of AI by a judge. In that case, the plaintiff challenged a judicial decision on the basis that it openly and heavily relied on ChatGPT. The Court found that the use of AI to draft a judicial decision does not violate, by its own, due process rights, as long as the tool does not replace judicial decision-making. In the case at issue, AI merely supported the reasoning after the judge had already reached a decision. Nonetheless, the Constitutional Court held that the the judge should have been more transparent about the way he used AI. The Court called on the Superior Council of the Judiciary to establish guidelines and, in the same ruling, set out guiding principles and criteria for the use of AI in the judiciary, including: (i) transparency; (ii) accountability; (iii) privacy; (iv) non-substitution of human rationality; (v) seriousness and verification; (vi) risk prevention; (vii) equality and equity; (viii) human control; (ix) ethical regulation; (x) compliance with good practices and collective standards; (xi) continuous monitoring and adaptation; and (xii) suitability.

In a February 2023 decision, the Administrative Tribunal of Magdalena authorised the holding of an initial hearing in a virtual environment resembling a Metaverse. In doing so, the tribunal noted that generative AI tools such as ChatGPT had been used in an orientative and transparent manner to assess the technical and legal feasibility of the hearing. The case illustrates that the use of AI as a supporting tool for logistical or organisational purposes, rather than for adjudicative reasoning, does not raise due process concerns when properly disclosed and limited in scope.

A related issue was addressed by the Supreme Court of Justice in decision STC17832-2025 (5 November 2025). In that case, the Supreme Court set aside a decision of the Superior Tribunal of Sincelejo that had terminated an enforcement proceeding on the basis of alleged jurisprudential precedents that were incorrectly attributed to the Supreme Court itself. Upon review, the Supreme Court found that the passages cited by the tribunal did not appear in the original decisions and that the ruling therefore relied on inaccurate or non-existent jurisprudential references, resulting in a defect of ‘apparent reasoning’ and a violation of the affected party’s due process rights.

In its reasoning, the Supreme Court emphasised that the duty to provide reasons for judicial decisions requires that the legal authorities invoked be authentic, precise and verifiable, since the legitimacy of judicial decisions depends on the transparency and accuracy of the legal sources relied upon. Referring to the Constitutional Court’s judgment T-323-2024, the Supreme Court also warned that the increasing use of digital tools and artificial intelligence in legal work makes it particularly important for judges and legal professionals to verify the reliability of the sources they cite.

Related issues were examined in STP21973‑2025, where the Supreme Court of Justice identified a serious violation of judicial impartiality after discovering that a draft conviction decision had been prepared and signed before the closing arguments were presented by the parties. The Court warned against the use of templates or automated drafting tools that anticipate judicial outcomes before procedural stages are completed, holding that such practices undermine the principle of the natural judge and the integrity of the adjudicative process. The Court annulled the proceedings and ordered the case to be reassigned to a different judge.

A similar concern arose in a criminal case decided by the Superior Tribunal of Bogotá (Auto of 2 December 2025, Rad. 2014‑00222), in which a conviction was declared absolutely null after it was found that the sentencing judge relied on legal citations to judgments that did not exist. The decision identified a failure of judicial motivation and a de facto delegation of the adjudicative function, noting that the drafting process relied on automated or AI‑assisted text generation that produced fictitious jurisprudential references. The tribunal stressed that judicial reasoning cannot be automated to the point of substituting verification of legal sources and ordered disciplinary referrals.

Algorithmic transparency and disclosure

In another ruling (T-067/2025), the Constitutional Court held in February 2025 that by refusing to disclose the source code of a public health application used to real-time monitoring of people’s health during the Covid pandemic (‘CoronApp’), the authorities violated the claimant’s fundamental right of access to public information as protected by the Colombian Constitution and Law 1712 of 2014 (Law of Transparency and Access to Public Information). The Court ordered the Ministry of Health to provide the source code and its version history, after taking necessary technical measures to protect personal data. It further instructed the government to develop guidelines for algorithmic transparency in public-sector digital tools.

The Court recognised that the source code of a public application such as CoronApp constitutes public information, particularly when the application is used to implement public policy and affects citizens’ rights. The Court emphasised the importance of algorithmic transparency in the public administration and observed that public access to the source code enables social oversight, helps detect potential biases or errors, and builds trust in public digital tools. It held that exceptions to such access must be interpreted restrictively. Any entity denying access must identify the precise legal basis for the exception and provide a detailed, evidence-based justification showing a real, probable, and specific harm that outweighs the public interest in disclosure. Notably, the Court clarified that copyright does not constitute a valid ground to deny access to public information held by public entities. In reaching its conclusion, the Court referenced international standards and comparative practices, noting that many jurisdictions published the source code of similar COVID-19 tracking applications to promote transparency and public confidence.

The decision reinforces the principle of transparency in the use of AI tools and algorithms within the public administration, which will likely be relevant to the use of AI in the criminal context.

Misuse of AI in Court filings

Recent judicial decisions have also addressed the improper use of AI by lawyers in judicial proceedings. In particular, in Auto AC739-2026 (13 February 2026), the Civil Chamber of the Supreme Court of Justice sanctioned a lawyer who submitted a petition containing multiple legal authorities in which sources could not be verified, including references to non-existent Supreme Court decisions and incorrect statutory provisions.

When asked to explain the inconsistencies, the lawyer acknowledged that the errors likely originated from the use of an artificial intelligence tool employed to draft the filing. The Supreme Court held that although the use of AI in legal practice is not prohibited, lawyers remain fully responsible for the accuracy of the legal sources cited in their submissions. It emphasised that the use of artificial intelligence does not relieve them of their professional duty of diligence and verification, even when AI tools are used to assist in drafting.

The Supreme Court found that the inclusion of several apocryphal citations (including references to non-existent legal provisions and jurisprudence, as well as citations to real decisions whose content did not correspond to what was attributed to them), which persisted even after the lawyer had been alerted to the irregularities, constituted procedural temerity. As a result, the Supreme Court imposed a fine equivalent to fifteen monthly minimum wages and referred the matter to the Judicial Disciplinary Commission for further review of potential professional misconduct.

A related example appears in AP760-2026, where the Criminal Chamber of the Supreme Court of Justice declared inadmissible a cassation appeal after concluding that the filing had largely been generated using artificial intelligence tools. The Supreme Court noted that the submission reproduced general argumentative suggestions produced by an AI system rather than properly formulated cassation grounds. Because the document failed to satisfy the technical requirements governing the extraordinary remedy of cassation, the Supreme Court rejected the appeal. Although no disciplinary sanction was imposed, the decision illustrates the risks associated with submitting AI-generated legal arguments without adequate verification and professional review. A more detailed discussion of this decision appears in the Defence section above.

Courts have also rejected attempts to introduce AI systems themselves as evidentiary tools. In a decision issued in March 2026 by the Civil Chamber of the Superior Tribunal of the Judicial District of Cali (Radicado No. 76‑001‑31‑03‑008‑2023‑00322‑01), the tribunal confirmed the refusal to admit a request whereby litigants sought to have “ChatGPT Pro” effectively participate in a hearing as a form of technical evidence. The purpose of the request was for the AI system to analyse medical records and expert reports in a medical liability case. The tribunal held that such a request was procedurally indeterminate and incompatible with orthodox evidentiary methodology, stressing concerns related to algorithmic opacity, the absence of an identifiable expert author accountable for the output, and the lack of a verifiable and reproducible technical method consistent with the rules governing expert evidence. The decision reinforces the limits of AI use within evidentiary proceedings and clarifies that generative AI systems cannot be treated as autonomous technical experts within judicial proceedings.

Deepfakes and synthetic media

As at March 2026, there are no reported instances of deepfake-generated evidence being adduced in criminal proceedings.