Introduction

Artificial Intelligence (AI) is undoubtedly the technological invention that is profoundly transforming society in this second decade of the 21st century. It is likely that there is no branch of knowledge or productive activity in society that is not currently reflecting on the use of artificial intelligence and the consequences it may have on tasks traditionally performed by humans.

The history of AI dates back decades, to when renowned scientist Alan Turing developed a machine during World War II that could think faster than humans and decrypt messages from the Nazi machine, Enigma, which encoded messages for German submarines targeting British ships. ³ In 1950, Turing published Computing Machinery and Intelligence, a landmark work introducing the Turing Test for AI. ⁴

Since the 1950s, we have witnessed tremendous advancements in AI, catalysed by the computing revolution of the 1970s and 1980s and the advent of machine learning in the 1990s. Projects like IBM’s Deep Blue, which defeated chess champion Garry Kasparov in 1997, ⁵ showcased AI’s potential. This technology laid the foundation for major computing advancements, including Google, which became the king of search engines and one of the most important technology companies in the world. By the 2010s, deep learning and neural networks emerged as the driving forces behind current AI breakthroughs. ⁶

Despite all the technological progress that preceded it, artificial intelligence gained significant global popularity starting in late 2022 with the launch of groundbreaking generative AI projects, such as OpenAI’s ChatGPT, funded by Microsoft. This project became so popular that it reached over 100 million active users in just two months. Other notable projects soon followed, including DALL-E, Bard, AlphaCode, and MidJourney, to name a few of the most representative. ⁷

As a result, the use of artificial intelligence has become highly significant across all areas of knowledge. ⁸ In commercial arbitration, both domestic and international, AI is significantly transforming how dispute resolution processes are conducted. This disruptive technology, applied across the various phases of arbitration, is enhancing efficiency, reducing costs, and fostering greater transparency and consistency in decisions. However, it also raises ethical and legal dilemmas that are worth considering. This article explores AI’s main uses in national and international arbitration and their effects on legal practice.

Early Applications of Artificial Intelligence in Arbitration

The application of artificial intelligence (AI) in arbitration has been a progressive development that has accelerated over the past two decades as technology has advanced and digital tools have gained acceptance in the legal field. ⁹ The use of AI in arbitration began very modestly in the 1990s when the technology was still in its early stages. Early advances in AI applied to arbitration were more related to the basic automation of repetitive tasks, such as document management and the analysis of legal texts. This period saw a significant increase in the use of massive databases like Lexis Nexis, Westlaw, Kluwer, and others. ¹⁰

At that time, expert systems—precursors to modern AI—began to be used for decision-making in certain areas of law, albeit in a very rudimentary way. Lawyers and arbitrators started experimenting with computerized databases to streamline the search for precedents and relevant legislation. One early example of such tools was software designed to classify legal documents and offer suggestions to help arbitrators manage large volumes of information. At this stage, AI functioned as an assistant for data collection, but key decisions remained in human hands.

In the 2000s, online dispute resolution (ODR) became one of the first areas where AI was introduced in arbitration. ODR platforms, such as the one eBay implemented to resolve disputes between its users, were pioneers in applying technology to resolve conflicts remotely, without requiring parties to meet in person. ¹¹

In these early systems, AI was primarily focused on mediation and facilitating agreements between parties, using simple algorithms to provide automated suggestions for resolution rather than playing an active role in making binding decisions. ¹²

With the rise of digitalization and big data, AI began to be applied in arbitration in more sophisticated ways. During this period, document management platforms and electronic discovery (e-discovery), which are used to provide digital evidence in arbitration cases, became essential tools for arbitration professionals. ¹³

As digitalization and big data surged, AI’s application in arbitration became more sophisticated. Tools for e-discovery, which facilitate evidence handling in arbitration, became essential for legal practitioners. AI-powered predictive coding tools revolutionized e-discovery by enabling efficient document review. For example, in Pyrrho Investments Ltd. v. MWB Property Ltd., the United Kingdom courts permitted predictive coding for document classification, ¹⁴ following protocols agreed upon by the parties, emphasizing proportionality in technology costs and case-by-case evaluation. ¹⁵

Legal firms began adopting AI to review the massive volumes of data and documents involved in arbitration processes. These technologies enabled faster classification, labeling, and analysis of documents compared to traditional methods. Machine learning became increasingly common, allowing systems to “learn” from data patterns and improve their accuracy in identifying relevant information. ¹⁶

From 2015 onwards, AI applications in arbitration advanced significantly with predictive analytics and algorithms capable of forecasting case outcomes. Platforms like Premonition ¹⁷ analyse vast datasets to predict arbitration results, processing over 174 million cases. ¹⁸

Similarly, Lex Machina provides predictive analytics based on historical data, helping parties estimate the likelihood of success in arbitration. This tool uses AI to analyze arbitrators’ behavior, especially in the United States, allowing parties to gain insights into a specific arbitrator’s past decisions, which can influence the selection of arbitrators and case strategy. Lex Machina employs AI and technology-assisted review to provide comprehensive case resolutions, damage analyses, solutions, findings, and accurate data about lawyers and parties. ¹⁹

As we will see later, AI has also been used to improve the arbitrator selection process. Systems can now analyze thousands of arbitrators’ prior decisions and professional profiles, providing parties with tools to determine potential outcomes in arbitration disputes.

Thus, the history of artificial intelligence (AI) applied to arbitration has evolved over the past decades in parallel with the development of disruptive technologies in the legal field. As AI tools have advanced, their integration into arbitration has shifted from an experimental concept to a practical reality, significantly impacting the efficiency of arbitration cases.

Uses of Artificial Intelligence in Arbitration

One of the most significant uses of AI in arbitration is in the contract analysis and risk management stage, where AI can review large volumes of commercial contracts. AI-based tools like Kira Systems exemplify how digital transformation is becoming imperative in the legal field. As law firms face cost-focused landscapes where client discounts are common, improving efficiency and creating a client-centred model is crucial.

This system has demonstrated that law firms using Kira AI reduce contract review times by up to 60 %. Leading firms such as DL Piper, Freshfields, and Allen & Overy have increased efficiency and enhanced client services through AI-powered software. ²⁰ This advantage is particularly vital in arbitrations requiring analysis of large volumes of contracts, giving a competitive edge to parties with access to such AI tools. Consequently, a case could be won not only by having the best arbitration litigators but also by leveraging the best AI tools in arbitration. ²¹

Similarly, platforms like Luminance allow parties to identify potentially conflicting or ambiguous clauses that could lead to disputes, facilitating a more thorough and quicker review process. Predictive analysis reduces litigation risks by helping companies identify areas of conflict before they escalate. Founded by AI experts from the University of Cambridge, Luminance is a leading legal AI platform. Based on a proprietary large legal language model (LLM), Luminance conceptually understands legal documents in any language. ²²

Luminance enhances and accelerates a wide range of tasks, from initial review of incoming contracts to automatic marking of contractual anomalies, highlighting areas of non-compliance, tagging clauses, and applying AI-driven Early Case Assessment (ECA) and Technology-Assisted Review (TAR) in evidence discovery matters. This platform is used by major law firms like Dentons and renowned companies such as Avianca and the BBC in London.

Another use lies in outcome prediction and risk assessment, as mentioned earlier. Tools like JuriMetría, developed by La Ley and Google Spain, and Arbilex, utilize extensive datasets of arbitration cases to forecast potential outcomes in future disputes. Machine learning algorithms identify patterns in previous decisions, enabling lawyers and parties to estimate success probabilities.

JuriMetría is a web platform that helps jurists analyse key aspects of judicial processes, such as duration, likelihood of appeals, and outcome predictions. It also provides insights into judges’ trajectories, jurisprudential trends, and arguments with higher success probabilities. With interactive graphs and access to legal rulings, it facilitates procedural strategy preparation and saves time in preliminary research. By combining advanced AI with La Ley’s legal expertise, JuriMetría optimizes legal work and improves client advisory accuracy. ²³

These AI tools not only aid in procedural strategy but also in negotiating agreements, promoting early resolutions, and reducing arbitration time and costs.

Document Review in Arbitration

Document review processes in arbitration, especially in international cases involving large volumes of evidence, can be lengthy and expensive. AI has revolutionized this area with Predictive Document Review (TAR). Programs like Relativity or Everlaw analyze millions of documents, identify patterns, and classify their relevance, accelerating the discovery of essential evidence for cases. Everlaw optimizes all aspects of litigation processes with features such as rapid data processing, email searches, and advanced analytics integrated into its platform. ²⁴

For example, in an international arbitration between two major tech corporations, AI could review over a million emails and documents in just a few days—tasks that would otherwise take months for a human team. This reduces costs and allows parties to focus on critical case aspects.

AI also plays a role in drafting awards. While final decisions remain human, AI can assist in structuring awards and researching relevant norms or precedents. Tools like ROSS Intelligence have been used by arbitrators and courts to find case law and simplify drafting complex arguments. ²⁵

Challenges and Legal Implications

However, the use of artificial intelligence systems that analyze large volumes of data to make such predictions also raises other legal issues, such as the permission to use such data and its relationship to copyright, ²⁶ particularly regarding access to content that is exclusive to its owners and is used without authorization by AI systems.

This is precisely the case of Ross Intelligence, which was sued by Thomson Reuters, the owner of Westlaw, for copyright infringement. In the first instance, Judge Stephanos Bibas, from the U.S. Court of Appeals for the Third Circuit, acting as a designated judge in the U.S. District Court for the District of Delaware, issued a memorandum mostly denying both parties summary judgment motions and sending the case to trial for a jury to decide on the disputed copyright infringement issues. ²⁷

The latest activity in the case dates back to July 2024, when the Court ordered ROSS to submit a list of all headings it considers to be direct quotations or that minimally vary from direct quotations of judicial opinions by July 29, 2024. Subsequently, if Reuters wished to submit comments on the list, it was required to do so by August 5, 2024.

The Court will not exclude any headings that are not direct quotations or trivial variations of direct quotations, meaning the jury will need to consider all headings that can reasonably be deemed original and, therefore, subject to copyright protection by the plaintiff. ²⁸

This case highlights the legal and ethical implications of using artificial intelligence as a tool for arbitrators and litigants in arbitration, but most importantly, for the designers of such tools.

In the international context, the growth of online arbitration has been largely driven by AI. Platforms such as Modria and Smartsettle allow parties to resolve commercial disputes entirely digitally. These platforms not only automate logistical aspects, such as document submission and communication management, but can also include AI algorithms that assist parties in finding solutions through AI-assisted negotiation.

Smartsettle is a highly secure negotiation system in the legal world that uses patented optimization algorithms to achieve fair and efficient solutions that go beyond the typical “win-win” outcomes for parties. Its founding company, iCan, invented a neutral-site negotiation support system that gathers and manages the preferences of any number of parties with conflicting goals on any number of issues, generating potential agreements based on the parties preferences. ²⁹

iCan’s intellectual property also includes the multivariate visual blind-bidding method and various other negotiation processes involving the use of artificial intelligence for dispute resolution.

In the field of document management, the development of AI-based tools such as ChatDoc, Sharly, and Humata is transforming this essential task for case analysis in various areas, including international arbitration. Their ability to efficiently analyze, summarize, and interact with documents makes them strategic allies in this field, where cases often involve large volumes of information and complex documentation. ³⁰

ChatDoc is a platform designed to allow users to upload documents and make specific inquiries through an interactive chat. ³¹ In the context of international arbitration, this tool facilitates the review of contracts, terms of reference, and parties’ briefs. Moreover, its ability to identify specific clauses within lengthy and complex texts significantly reduces analysis time and improves accuracy in managing key documents.

Meanwhile, Sharly stands out for its ability to extract data and generate automated summaries. In an international arbitration process, this tool can be used to summarize lengthy documents such as arbitral awards, parties’ submissions, or expert reports. ³² Additionally, it helps identify patterns in international legal precedents, which supports the development of more robust and outcome-oriented legal strategies.

Humata takes document interaction to the next level by allowing users to ask specific questions about uploaded texts and receive direct and accurate answers. ³³ In international arbitration, this tool is especially useful for queries related to applicable regulations, international treaties, or arbitration rules. It is also a powerful tool for verifying the consistency of documentation submitted by the parties, ensuring that there are no inconsistencies or errors in legal arguments.

Together, these tools offer numerous benefits to international arbitration. They optimize document management by handling large volumes of information in an organized manner, reduce costs and time by automating repetitive tasks, and increase accuracy in case preparation. Additionally, by offering real-time analysis, they enable legal teams to make informed decisions more quickly. With their application, international arbitration faces a future that is more efficient, precise, and technologically advanced.

In the coming years, we will likely see how leading arbitration centers in our region, such as the Arbitration and Conciliation Center of the Bogotá Chamber of Commerce, will incorporate AI-assisted arbitration services, following the leadership it has always had by being one of the first centers in the region to offer online arbitration tools.

In international disputes, where language differences, time zones, and travel costs can complicate the arbitration process, AI-based platforms have enabled faster and more accessible resolutions. One example is the case of international e-commerce, where companies disputing product quality or contractual terms have been able to resolve their conflicts in weeks rather than months, using online resolution platforms that employ AI to facilitate negotiation. ³⁴

The use of artificial intelligence (AI) in arbitration practice is rapidly transforming global processes, becoming an indispensable tool in international arbitration. A detailed analysis of the Annual Report Survey, The Rising of Machine Learning, prepared by Bryan Cave Leighton Paisner, ³⁵ reveals the breadth and implications of AI use in this sector. The study highlights that 90 % of professionals surveyed are familiar with AI tools applied to arbitration, demonstrating its penetration in legal practice. Specifically, 37 % use AI for document translation, 30 % for text review and formatting, and 24 % for complex document analysis—activities that traditionally required significant human and financial resources.

However, this technological advancement is not without controversy. ³⁶ Despite the evident advantages, such as the significant time savings identified by 85 % of those surveyed, concerns about reliability and ethics in the use of these tools arise. 88 % of participants expressed concern about the possibility that AI could generate fictitious or erroneous information, questioning the trustworthiness of its results. ³⁷ These concerns are amplified when considering the use of AI in more complex and sensitive tasks, such as drafting legal texts. In fact, 53 % of respondents believe that AI-generated legal content should be limited, reflecting a preference for maintaining human control over critical aspects of arbitration. ³⁸

Moreover, the ethical and regulatory debate surrounding AI is becoming increasingly prominent. 60 % of respondents demand greater transparency in how these technologies are implemented, while 74 % oppose the use of AI by arbitrators to draft awards. This reflects a widespread perception that automation in such delicate decisions could compromise the legitimacy and fairness of the arbitration process. ³⁹ Lastly, 63 % support the creation of specific regulatory frameworks for the use of AI, a clear signal that, although its potential is undeniable, trust in these tools is conditional upon the existence of a robust system of controls and guarantees, as we will discuss later.

In this context, it is evident that the path toward a balanced integration of AI in arbitration requires a multidisciplinary approach that combines technological advances with solid ethical and legal principles. Only in this way will it be possible to maximize the capabilities of these tools while mitigating the risks inherent in their use. ⁴⁰

In line with this, a survey conducted by the prestigious Queen Mary London University School of International Arbitration and White & Case in 2021 found that the use of artificial intelligence in arbitration practice is becoming more frequent, with 41 % of respondents using AI always, frequently, or sometimes in their arbitration practice. ⁴¹

Interviews conducted in the survey showed that the use of AI in the legal field is almost exclusively concentrated on technology-assisted document review, known as e-discovery. This approach has been described by one of the interviewees as a “brilliant revolution,” as it has significantly improved procedural efficiency. The ability to process and analyse large volumes of data in shorter timeframes has allowed the parties involved in international arbitrations to save resources and time, which is crucial in alternative dispute resolution procedures. However, the potential of AI in other areas is still being explored, with few consolidated cases beyond this specific application. ⁴²

In addition to e-discovery, some interviewees mentioned the occasional use of AI-based data analysis tools, which can help identify relevant patterns in complex litigation or arbitration cases. However, a recurring issue is the high cost of these tools, making them inaccessible to many users, especially in disputes of lower economic value or less technical complexity.

This economic aspect becomes a barrier, as even when clients have the resources to invest in these technologies, they often question whether the benefits justify the expense. This is particularly true when disputes do not involve large volumes of information or require advanced analysis. Additionally, the free tools available today carry the risk of “hallucinating” or providing incorrect information, to the point of inventing cases.

Such was the case on March 8, 2024, when the Federal Court for the Middle District of Florida suspended a lawyer for one year after discovering that they had submitted a brief containing non-existent case law precedents generated by artificial intelligence. This violated both the court’s rules and the Florida Bar’s Rules of Professional Conduct.

The case began when the opposing counsel filed a complaint stating that the cited cases could not be found in any legal database. Upon requesting full versions of these precedents, they received evasive responses from the sanctioned lawyer, who ultimately admitted to using artificial intelligence to draft parts of the document without verifying the accuracy of the generated citations. This prompted the Grievance Committee to investigate further, concluding that the lawyer’s conduct demonstrated not only a lack of reasonable diligence but also the invocation of completely fabricated precedents. ⁴³

In its report, the Committee emphasized that while artificial intelligence can be a useful tool in legal research, it does not replace the lawyer’s responsibility to ensure the accuracy and validity of the legal arguments presented to the court. As a result, Judge John E. Steele, based on clear and convincing evidence, suspended the lawyer and imposed strict conditions for reinstatement. These included attending professionalism and management workshops, complying with corrective measures ordered by the Florida Bar and other courts, settling pending sanctions, and certifying the complete reading of the rules of professional conduct and the court’s local rules. ⁴⁴

This case underscores the importance of professional diligence in the practice of law, even when using advanced technological tools like artificial intelligence. The sanction sends a clear message: lawyers have an unyielding obligation to verify the accuracy of the information they present in their legal briefs.

Another challenge identified is the lack of familiarity with new technologies among arbitration users. The implementation of AI in this field not only depends on costs but also on the acceptance and understanding of professionals and involved parties. In some cases, there remains a sense of distrust toward these tools, especially when it comes to algorithms designed to predict legal outcomes or judicial decisions, a practice known as predictive justice.

Many interviewees in the Queen Mary and White & Case report expressed scepticism regarding the ethics and limits of these tools, questioning to what extent they should interfere with the adjudicative functions of human judges and arbitrators.

Despite the advantages AI can offer, such as improving efficiency and reducing human errors, there is a general consensus that these tools cannot replace the judgment and experience of arbitrators and lawyers. Human interaction remains indispensable for interpreting the complexities of each case and ensuring a fair and ethical approach to conflict resolution. Indeed, most interviewees emphasized that AI should be considered a complementary tool, not a substitute, as the human element is essential to ensure that legal decisions are not only technically correct but also socially just and acceptable. ⁴⁵

Finally, the interviews also highlighted the need to foster greater technological literacy in the legal field. This involves not only training professionals in the use of these tools but also addressing ethical concerns and the biases inherent in algorithms, which could negatively impact trust in the justice system.

The successful implementation of AI in arbitration and other legal areas will require a balanced approach that combines technological innovation with a strong commitment to the fundamental values of the law. Below is the chart referenced in the Queen Mary and White & Case report on the use of AI in international arbitration (Figure 1).

The previous cases and situations identified by arbitrators and arbitration practitioners worldwide highlight that, despite the rapid growth of these important technological innovations applied to both domestic and international arbitration, one of the most significant we face as arbitrators and developers of arbitration technology is ensuring that the use of AI in arbitration maintains one of the justice system’s most cherished values: impartiality. ⁴⁶

Regulation of Artificial Intelligence Applied to Arbitration

AI algorithms can replicate biases present in the data used to train them, potentially influencing the prediction of outcomes or pattern identification. Therefore, it is essential for developers and users of AI in arbitration to adhere to high ethical standards and maintain transparency regarding the data used to train these systems.

This concern has already led to the development of AI regulations, such as the European Union’s Artificial Intelligence Act, approved in 2024. It is the first comprehensive regulation in the world addressing the risks and opportunities of this technology. Its primary goal is to ensure that AI systems are safe, respect the fundamental rights of European citizens, and foster technological innovation within an ethical framework. ⁴⁷

The regulation classifies AI systems according to their level of risk, ranging from low to unacceptable. It sets specific requirements for high-risk systems, such as those used in critical areas like education, healthcare, banking, or border control. Moreover, it bans practices deemed unacceptable, such as mass facial recognition without consent, biometric categorization based on sensitive characteristics, and systems that manipulate human behavior ⁴⁸ or exploit vulnerabilities, which could potentially be used in judicial or arbitral decisions.

The law also mandates that AI systems be transparent, accurate, and subject to human oversight. It requires disclosing if content has been created or modified using AI and establishes significant penalties for non-compliance. While imposing clear restrictions, the regulation seeks to balance the protection of fundamental rights with encouraging investment and technological development in Europe. This legal framework, applicable to both companies within and outside the EU whose systems are used within European territory, sets a global precedent for the responsible management of emerging technologies. ⁴⁹

Similarly, arbitration regulations are beginning to address the use of AI in the arbitration process. The Silicon Valley Arbitration and Mediation Center (SVAMC) pioneered this field with its 2024 Guidelines on the Use of Artificial Intelligence in Arbitration. This document aims to integrate emerging technologies into arbitration, promoting greater efficiency, accessibility, and transparency while ensuring the protection of the rights of all parties involved. ⁵⁰

The guidelines focus on providing a practical, technical, and ethical framework for adopting AI at each stage of the arbitration process.

First, the guidelines emphasize the criteria for the responsible selection and use of AI tools. They recommend that parties adopt technologies that guarantee impartiality, security, and privacy in managing information. Additionally, they highlight the importance of continually validating these tools to ensure compliance with appropriate technical and legal standards. This includes using AI for tasks such as predictive analysis and data management, which can significantly reduce the time and costs of arbitration.

However, the guidelines place significant emphasis on the arbitrator’s responsibility for the decisions they make. A key principle highlighted is that an arbitrator’s mandate—especially the decision-making function—is personal and non-delegable. The guidelines do not prohibit the use of AI tools to assist arbitrators in analysing facts, arguments, evidence, ⁵¹ and the law or drafting a reasoned decision. ⁵²

While AI tools can handle information, analyse data, and predict outcomes, they must not replace the human judgment, discretion, accountability, and responsibility inherent in the arbitrator’s role. Arbitrators must ensure that they do not inadvertently delegate part of this personal mandate to AI tools.

The guidelines reinforce that arbitrators must review the results generated by any AI tool to ensure their accuracy and take responsibility for any errors or inaccuracies. If an arbitrator uses a generative AI tool to assist in analysing arguments or drafting part of a decision or award, they cannot simply reproduce the AI’s output without ensuring it appropriately reflects their personal and independent analysis of the issues and evidence.

Finally, the guidelines remind arbitrators that, despite technological advancements, their personal responsibility in decision-making remains fundamental. While AI tools can enhance efficiency and provide insights, the arbitrator must make the final decision, preserving the essential human element required for fairness and integrity in the arbitration process. The arbitral tribunal is not exempt from personally reviewing the case file and/or drafting any tribunal decisions. At all times, arbitrators remain accountable for the use of AI during arbitration. ⁵³

Another key point is the role of AI in case management. The guidelines address how AI tools can facilitate the organization of documents, the analysis of large volumes of information, and the automation of administrative processes. Despite these advantages, the SVAMC stresses the need for substantive decisions to remain under human oversight to preserve fairness and balance in arbitration. ⁵⁴

In terms of transparency and ethics, the guidelines establish principles to ensure that AI tools are explainable and understandable to all parties. This includes identifying and mitigating potential biases in algorithms that could affect arbitral decisions. Additionally, safeguards are recommended to protect the confidentiality of information and prevent security breaches when using advanced technologies. ⁵⁵

Finally, the SVAMC recognizes that training is essential for the successful adoption of AI in arbitration. Arbitrators, lawyers, and parties are encouraged to receive education on the operation, capabilities, and limitations of these tools, fostering a comprehensive understanding of their impact on proceedings.

In conclusion, the SVAMC guidelines offer an innovative vision for the future of arbitration, highlighting the potential benefits of AI while addressing ethical and practical challenges. Although AI promises to revolutionize arbitration by optimizing resources and improving processes, the guidelines emphasize the importance of maintaining human oversight to ensure fairness and justice in every decision.

Most recently, on October 16, 2024, the Arbitration Institute of the Stockholm Chamber of Commerce also published guidelines on the use of artificial intelligence (AI) in cases administered under its rules. ⁵⁶ AI, a rapidly evolving field, has the potential to enhance efficiency and reduce costs in arbitration, which has been criticized for increasing time and expenses in nearly all jurisdictions worldwide. ⁵⁷

The guideline, designed to be flexible and updatable, aims to provide direction to arbitration participants without imposing obligations. It will only apply when it does not conflict with agreements or mandatory arbitration rules and public policy at arbitration venues.

AI is defined as systems that, with varying levels of autonomy, can adapt and generate predictions or decisions based on data. ⁵⁸ In arbitration, as discussed in this article, AI is used for tasks such as legal research, evidence review, translation, transcription, drafting communications, predictive decisions, and analysis of large volumes of documents and evidence.

The use of AI in arbitration holds great potential for reducing costs and increasing efficiency, but it also poses risks. The guidelines highlight several considerations, aligning with the SVAMC directives, as follows:

• Confidentiality: Participants are advised to understand how data is managed when using AI.

• Quality: AI can perpetuate biases or generate incorrect information, making human review essential to ensure the quality of decisions.

• Integrity: Transparency and accountability are crucial, and arbitral tribunals must disclose the use of AI in research and decision-making.

• Non-delegation: AI tools should support, not replace, the decisions of arbitral tribunals. ⁵⁹

As AI usage increases, concerns about ethics and regulation are emerging and are being addressed by arbitration centers worldwide. Discussions on ensuring these technologies are used fairly and equitably are ongoing and will likely shape the regulations of leading arbitral institutions globally. This underscores the importance of closely monitoring regulations on the topic, particularly those currently under consideration in Colombia’s Congress, as they will undoubtedly impact arbitration. ⁶⁰

Conclusions

The journey of artificial intelligence (AI) in arbitration reflects a progression from early experimental applications to the integration of advanced technologies that significantly enhance efficiency and effectiveness in resolving disputes. Over time, these innovations have transformed arbitration, positioning it as a domain uniquely capable of leveraging AI to address the complexities of modern conflict resolution. Yet, this transformation also demands a critical examination of the ethical, legal, and regulatory challenges that arise alongside the benefits.

The adoption of AI in arbitration has proven instrumental in accelerating processes, increasing precision, and improving decision-making accuracy. In an era of globalization and interconnected economies, such advancements are crucial for managing the growing volume and complexity of disputes. AI tools, such as natural language processing, predictive analytics, and automated document review, have been particularly impactful in reducing the time and resources required for case preparation and evaluation, thereby enhancing overall transparency and fairness.

Despite these advantages, the integration of AI into arbitration is not without significant challenges. Ethical considerations, including biases inherent in algorithms, data privacy concerns, and the potential for over-reliance on technology, underscore the need for robust governance frameworks. Human oversight remains critical to ensure that these tools are used responsibly and that their outputs align with the principles of justice and equity. Proper regulation and supervision will be indispensable in mitigating risks and fostering trust in AI-driven arbitration processes.

Recent years have demonstrated the growing importance of AI in both national and international arbitration contexts. For instance, technology has played a pivotal role in managing large datasets, automating repetitive tasks, and improving the accessibility of arbitration proceedings. These applications have already showcased AI’s potential to optimize the resolution of complex disputes, saving valuable time and resources.

Looking ahead, it is anticipated that AI will become an even more integral component of arbitration, evolving from a supplementary tool to a foundational element in dispute resolution. However, realizing this potential will require a thoughtful approach to balancing technological innovation with human judgment. Striking this balance is essential not only to maximize the benefits of AI but also to preserve the fundamental human-centric values that underpin the arbitration process. As the field continues to evolve, ongoing dialogue among stakeholders—arbitrators, technologists, policymakers, and parties—will be vital to navigate this dynamic intersection of technology and law effectively.

References

Notes

About the Author

Research Professor of the GEPPI research group at the Catholic University of Colombia. Master’s in Intellectual Property from Queen Mary University of London. Arbitrator at the Arbitration and Conciliation Center of the Bogotá Chamber of Commerce. President of the Colombian Legal Tech Association (alt+co).

^* Research paper

¹ Juan Carlos Galindo Vácha, La democracia digital, 73 Vniversitas (2024).

² Galindo Vácha, supra note.

³ Antoni Vidal, Alan Turing y el Nacimiento de la Inteligencia Artificial, Antena de Telecomunicación 45-45 (2007).

⁴ Alan Turing, Computer Machinery and intelligence, 236 Mind A Quarterly Review Of Psychology And Philosophy, 433-460, (1950).

⁵ Quique Rayuela, La batalla entre el hombre y la máquina: Garry Kasparov contra Deep Blue (Principia Magazine, 2023).

⁶ Felipe Rosano, Fundamentos De Redes Neuronales Artificiales (UNAM, 2017).

⁷ Verónica Smink, Las 3 etapas de la Inteligencia Artificial: en cuál estamos y por qué muchos piensan que la tercera puede ser fatal (BBC Mundo, 2023).

⁸ Rubén Méndez Reátegui, National Development Plans and Market Process: A revision from Law & Economics, IA, and the Public Issues perspectives, 73 Vniversitas (2024).

⁹ Jairo Becerra, Germán Darío Flórez Acero , Claudia Bibiana García Vargas , Marco Emilio Sánchez Acevedo , Carolina Rojas Orjuela & Jheison Torres Ávila, El derecho y las tecnologías de la información y la comunicación (TIC) (Universidad Católica de Colombia, 2014).

¹⁰ Jorge Luis González, Arbitraje en distintas áreas del derecho (Biblioteca de Arbitraje, 2007).

¹¹ Nicolás Lozada, Sistemas y tecnología de inteligencia artificial en la resolución de controversias (Blog Derecho de los Negocios Universidad Externado de Colombia, 2021).

¹² Id.

¹³ Carlos Coscollano, Los Beneficios del E-discovery en las Investigaciones Forensic (BDO Spain, 2021).

¹⁴ High Court of Justice Chancery Division, England, Pyrrho Investments Ltd. v. MWB Property Ltd, [2016] EWHC 256 (Ch) (2016).

¹⁵ Aditya Chauhan, Future of AI in Arbitration: The Fine Line Between Fiction and Reality (Kluwer arbitration blog, 2020).

¹⁶ Daniel Segura & Jaime Segura, Machine learning para arbitraje financiero en el mercado de renta variable colombiano (Universidad del Rosario, 2022).

¹⁷ Ben Davies, Arbitral Analytics: How Moneyball Based Litigation/Judicial Analytics Can Be Used to Predict Arbitration Claims and Outcomes, Vol. 22, No. 2 Pepperdine Dispute Resolution Law Journal 321-376 (2022).

¹⁸ Premonition AI, The World Largest Litigation Database (2024).

¹⁹ Lex Machina, How Lex Machina Legal Analytics Works (2024).

²⁰ Kira. (2023). How Law Firms Leverage Kira’s AI to Cut Contract Review Time by up to 60%.

²¹ Alberto Acevedo, Inteligencia artificial en el arbitraje internacional: de la predicción legal a los laudos emitidos por robots, (Garrigues, 2019).

²² Luminance AI. AI for the Spectrum of Legal Matters (2024).

²³ Jurimetría, La herramienta de analítica jurisprudencial predictiva más innovadora del mercado (La Ley, 2022).

²⁴ Everlaw, Collaborative Trial Preparation (Everlaw, 2024).

²⁵ Chauhan, supra note 14.

²⁶ O. Woolcott & G. D. Flórez, El régimen de exención de responsabilidad de los ISP por infracciones de propiedad intelectual en el TIC Colombia-estados unidos: una explicación a partir de la DMCA y la DCE, 63(129) Vniversitas 385-416 (2014).

²⁷ B. Ambrogi, Breaking: Judge Rules Ross Intelligence Copied Westlaw’s Headnotes, But Leaves for Jury to Decide if It Violated Thomson Reuters’ Copyright (Law Sites, 2023).

²⁸ Court of the District of Delaware, Thomson Reuters v. ROSS, Case 1:20-cv-00613-SB.

²⁹ Smartsettle, Beyond win (Smartsettle, 2022).

³⁰ Hernando Herrera Mercado, El arbitraje y la inteligencia artificial (III): aplicaciones útiles (Ámbito Jurídico, 21 de noviembre de 2024).

³¹ ChatDoc, Chat with documents, Get instant answers with cited sources (Chatdoc, 2024).

³² Sharly, Chat with documents. Summarize long documents and simplify complex PDFs with AI Chat (Sharly, 2024).

³³ Humata, AI meets your knowledge base, Ask me (Humata, 2024).

³⁴ Alejandro Morales Cáceres, El impacto de la inteligencia artificial en el Derecho, No. 039 Advocatus 39-71 (2021).

³⁵ Bryan Cave Leighton Paisner, Annual Arbitration Survey. The Rise of the machine learning (BCLP, 2023).

³⁶ Darío Parra Sepúlveda & Ricardo Concha Machuca, Inteligencia artificial y derecho. Problemas, desafíos y oportunidades, 70 Vniversitas (2021).

³⁷ Herrera Mercado, supra note 29.

³⁸ Leighton Paisner, supra note 34.

³⁹ CIAR Global, ¿Cómo está impactando la inteligencia artificial en el arbitraje internacional? (CIAR, 2023).

⁴⁰ Leighton Paisner, supra note 34.

⁴¹ Queen Mary University & White and Case, The 2021 International Arbitration Survey, Adapting Arbitration to a Changing World (Queen Mary University & White and Case, 2021).

⁴² Id.

⁴³ Carlos Fernández, Un Tribunal federal norteamericano suspende de ejercicio a un abogado por presentar un escrito procesal elaborado por IA que incluía precedentes inexistentes (La Ley, 25 de marzo de 2024).

⁴⁴ Bob Ambrogi, Federal Court Suspends Florida Attorney Over Filing Fabricated Cases Hallucinated by AI (LawSites, Mar. 14, 2024).

⁴⁵ Queen Mary University & White and Case, supra note 40.

⁴⁶ Sebastián Bozzo Hauri, El uso de las nuevas tecnologías como forma de disminuir las barreras de acceso a la justicia del consumidor en Chile, 72 Vniversitas (2023).

⁴⁷ Salvador García M., Derechos humanos y tecnologías de la información y la comunicación, 18(2) Novum Jus 225-248 (2024).

⁴⁸ European Union Artificial Intelligence Act 2024.

⁴⁹ Nuria Bigas, Cinco claves sobre la nueva Ley de Inteligencia Artificial de la Unión Europea (Universidad Abierta de Cataluña, 2024).

⁵⁰ Directrices sobre el uso de inteligencia artificial en el arbitraje del Sillicon Valley Arbitration and Mediation Center (SVAMC) de 2024.

⁵¹ Marcela María Yepes Gomez, Jesus Armando Perez Benitorevollo & Mario Peinado, Aplicación de la prueba electrónica en el marco normativo colombiano, 16 (1) Novum Jus 253-77 (2022).

⁵² Directrices sobre el uso de inteligencia artificial…, supra note 49.

⁵³ CIAR Global, Directrices de la SVAMC sobre uso de Inteligencia Artificial en arbitraje (CIAR, 2024).

⁵⁴ José Pablo Valverde, Disrupción digital de los medios de prueba (Consotium Legal, 2024).

⁵⁵ Directrices sobre el uso de inteligencia artificial…, supra note 49.

⁵⁶ Guide to the use of AI in cases administered under the SCC Rules 2024.

⁵⁷ José Fernández Rosas, Guía para el uso de la IA en casos administrados de conformidad con las Normas del Instituto de Arbitraje de la Cámara de Comercio de Estocolmo (Blog José Fernández Rosas, 16 de octubre de 2024).

⁵⁸ European Union Artificial Intelligence Act 2024.

⁵⁹ SCC Arbitration Institute, The SCC Arbitration Institute leads the way in embracing the age of AI (SCC Rules, 2024).

⁶⁰ Dalia Carreño Dueñas, Pensar el derecho como derecho virtual (Universidad Católica de Colombia, 2019).

Author notes

^aAutor de correspondencia / Correspondence author. E-mail: gdflorez@ucatolica.edu.co

Additional information

Cómo citar este artículo/How to cite this article: Germán Darío Flórez Acero, The Use of Artificial Intelligence in Arbitration: Friends with Benefits, 74 Vniversitas (2025), https://doi.org//10.11144/Javeriana.vj74.uaia

Descargar