Artificial Intelligence (“AI”) has made significant inroads across a series of sectors, as agriculture, e-commerce, education, entertainment, finance, healthcare, manufacturing, natural language processing, retail, transportation, security, and surveillance. But what about the legal sector? Since 1999, domain name disputes have been effectively solved under the Uniform Domain Name Dispute Resolution Policy (the “Policy” or “UDRP”) a procedure inspired in arbitration, which can be perceived as a predictable process. This article seeks to navigate into this question: What is the impact of AI in out-of-court procedures in the European Union? Is it possible to decide domain name disputes through the use of AI?

On April 30, 1999, the World Intellectual Property Organization (“WIPO”) published the Final Report of the First WIPO Internet Domain Name Process. Based on the proposal of the Government of the United States of America, and with the approval of WIPO Member States, from July 1998 WIPO lead an extensive international process of consultations seeking to make recommendations to the Internet Corporation for Assigned Names and Numbers (ICANN) on certain questions arising out of the interface between domain names and intellectual property rights. On that process 17 consultation meetings were held in 15 different cities throughout the world, and written submissions were received from 334 governments, intergovernmental organizations, professional associations, corporations, and individuals.

At its meetings on August 25 and 26, 1999 in Santiago, Chile, the ICANN Board of Directors adopted the UDRP Policy (the “Policy” or “UDRP”) based on the recommendations contained in the Final Report of the WIPO Internet Domain Name Process as well as comments submitted by registrars and other interested parties. Since then, all ICANN-accredited registrars that are authorized to register names in the generic Top-Level Domains (“gTLDs”) and the Country Code Top-Level Domains (“ccTLDs”) that have adopted the Policy have agreed to abide by and implement it for those domains, and any person or entity wishing to register a domain name in the gTLDs and ccTLDs in question is required to consent to the terms and conditions of the Policy.[1]

On October 24, 1999, the ICANN Board adopted a set of Rules for Uniform Domain Name Dispute Resolution Policy (the “Rules”) setting out the procedures and other requirements for each stage of the dispute resolution administrative procedure. The Policy procedure is administered by dispute resolution service providers accredited by ICANN.[2] Currently, the list of approved Dispute Resolution Service Providers is as follows: Asian Domain Name Dispute Resolution Centre (ADNDRC); Canadian International Internet Dispute Resolution Centre (CIIDRC); Czech Arbitration Court (CAC); National Arbitration Forum (NAF); and the leader, the WIPO Arbitration and Mediation Center (WIPO Center).

After 25 years, more than 80,000 domain name disputes have found an effective resolution through the Policy. The Policy is one of the most significant and reliable out-of-court solutions to any trademark owner. Facts that bring an inevitable question: Why has it been so successful? To this writer, several factors attribute to the success of the Policy along the years. As a number one, the Policy is the result of a profound international process, which implied effort, time and resources, conceived and conducted by an organization as WIPO; as a number two, the Policy has been able to provide a worldwide solution to trademark owners who seek a Decision from an impartial Expert (Domain Name Panelist); third, the process is mostly conducted online; fourth, each Service Provider based on the same Policy and Rules, through its own Supplemental Rules, expertise and geographic location, adds value to the domain name out-of-court system: as it is the case of the WIPO Center, which provides the leading reference in terms of the Domain Name Jurisprudence (WIPO Overview 3.0), or the CAC through its own Platform, or CIIDRC through its own Domain Name Educational Channel; and as the fifth factor, despite it might be considered as delusional to mentioned it, the entire process has been and is conducted and performed by us, the Humans.

An assertion that can make us wonder, if it is possible that four decades ago James Cemeron tried to alert about such potential dispute “Humans driven process v. Artificial Intelligence (‘AI’)”.[3]

The “Humics” [4]

According to Pascal Bornet, human beings own a series of core abilities that are timeless and transferable across various fields and industries, defined as “Humics,” meaning:

“Genuine creativity goes beyond the mere recombination of existing ideas that AI can perform. It involves the human capacity to conceive truly original concepts and paradigms, often driven by emotional resonance that machines cannot replicate.

Critical thinking encompasses our ability to solve unstructured problems, make ethical judgments, and question information in context – areas where AI falls short.

Social authenticity, including emotional intelligence and empathy, allows us to connect with others on a deep level, providing leadership and communication skills that remain distinctly human.”[5]

Also, according to Bornet, it is possible to anticipate the emergence of new skills that blend Humics + AI in unique ways:

“Emotional Innovation: Combining genuine creativity with social authenticity to create products or experiences that resonate deeply on a personal, human level.

Intuitive Complex Systems Navigation: Leveraging critical thinking and social authenticity to manage intricate systems involving both human and AI elements.

Human-AI Collaboration Leadership: Developing the ability to lead teams that seamlessly integrate human and AI capabilities, maximizing the strengths of both.”[6]

Human being’s set of emotions can reach destructive scenarios, our own History is a tangible proof of that, however, conversely, we as humans are in full capacity of being the authors of greater things in the name of real progress, or as one of my favorite pop rock bands use to proudly profess: “In the name of love”[7]. Therefore, how shall we deal with AI in the legal sector? Is it possible to be fully replaced by the AI? Is AI capable of solving domain name disputes by itself? Can AI actually be a legal entity? Could this even lead to an Electronic Panelist? What is the EU approach in relation to those questions?

EU AI Act

Alternative dispute resolution encompasses a range of processes, including mediation, arbitration, and negotiation, which aim to resolve disputes outside traditional court settings. ADR has gained prominence for its efficiency, cost-effectiveness, and ability to provide parties with more control over the resolution process. However, the integration of AI into ADR mechanisms raises significant concerns regarding fairness, accountability, and the potential erosion of human oversight.

Domain Name Dispute Resolution is considered as a form of Alternative Dispute Resolution (ADR). Therefore, it may be implicit, that the questions towards the Policy procedure, for those bodies located in the EU, are applicable.

According to Annex 3[8] referring to “High-Risk AI Systems Referred to in Article 6(2)”, section 8 and recital sixty-one (61) of the European Union’s Artificial Intelligence (“EU AI”) Act, which states that:

High-risk AI systems pursuant to Article 6(2) are the AI systems listed in any of the following areas:

8. Administration of justice and democratic processes:

(a) AI systems intended to be used by a judicial authority or on their behalf to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts, or to be used in a similar way in alternative dispute resolution; Related: Recital 61.

Recital sixty-one[9]

Certain AI systems intended for the administration of justice and democratic processes should be classified as high-risk, considering their potentially significant impact on democracy, the rule of law, individual freedoms as well as the right to an effective remedy and to a fair trial. In particular, to address the risks of potential biases, errors and opacity, it is appropriate to qualify as high-risk AI systems intended to be used by a judicial authority or on its behalf to assist judicial authorities in researching and interpreting facts and the law and in applying the law to a concrete set of facts. AI systems intended to be used by alternative dispute resolution bodies for those purposes should also be considered to be high-risk when the outcomes of the alternative dispute resolution proceedings produce legal effects for the parties. The use of AI tools can support the decision-making power of judges or judicial independence but should not replace it: the final decision-making must remain a human-driven activity. The classification of AI systems as high-risk should not, however, extend to AI systems intended for purely ancillary administrative activities that do not affect the actual administration of justice in individual cases, such as anonymization or pseudonymization of judicial decisions, documents or data, communication between personnel, administrative tasks. (Emphasis added).

Article 3: Definitions

(2) ‘risk’ means the combination of the probability of an occurrence of harm and the severity of that harm (Emphasis added).

The Implications of Recital 61 of the EU AI Act on Alternative Dispute Resolution: A High-Risk Classification

The EU AI Act, a landmark legislative proposal aimed at regulating AI technologies within the EU, has garnered significant attention for its comprehensive framework addressing various facets of AI deployment. Among its numerous provisions, Recital 61 specifically underscores the classification of AI applications in ADR as high-risk. This classification invites critical examination, as it reflects the EU’s broader commitment to safeguarding fundamental rights, ensuring transparency, and maintaining the integrity of legal processes in the face of rapidly evolving technologies.

High-Risk Classification in the EU AI Act

The EU AI Act categorizes AI systems into four risk levels: minimal, limited, high, and unacceptable risk. Recital sixty-one positions AI applications in ADR within the high-risk category, a designation that necessitates stringent regulatory requirements and oversight. The rationale behind this classification stems from several interrelated concerns:

1. Impact on Fundamental Rights: The deployment of AI in ADR processes can significantly affect individuals’ rights and interests. Given that disputes often involve sensitive issues, the use of AI tools without adequate safeguards could lead to biased outcomes, undermining the fairness of the resolution process. The EU’s commitment to upholding fundamental rights demands careful scrutiny of how AI technologies influence decision-making in ADR contexts.

2. Transparency and Explainability: AI systems, particularly those utilizing complex algorithms and machine learning techniques, often operate as ‘black boxes’.[10] This lack of transparency poses challenges in understanding how decisions are made, which is particularly problematic in legal contexts where parties must comprehend the rationale behind outcomes. Recital sixty-one emphasizes the need for transparency and explainability in AI systems used in ADR to ensure that parties can challenge and understand decisions.

3. Accountability: The high-risk designation imposes an obligation on providers of AI systems to ensure accountability mechanisms are in place. This includes robust governance structures that facilitate monitoring and auditing of AI decision-making processes. In ADR, where the stakes can be high, ensuring that there are clear lines of accountability is crucial to maintain trust in the system.

4. Potential for Bias and Discrimination: AI systems are susceptible to perpetuating existing biases present in training data. In the context of ADR, this risk is particularly acute as biased algorithms could lead to discriminatory outcomes based on race, gender, socioeconomic status, or other protected characteristics. Recital sixty-one highlights the necessity for rigorous testing and validation of AI systems to mitigate such risks.

5. Real Human Supervision: While AI can enhance efficiency in dispute resolution, the importance of real human supervision cannot be overstated. Recital sixty-one advocates for maintaining human involvement in decision-making processes, ensuring that AI serves as a tool to assist rather than replace human judgment. This aligns with the EU’s broader commitment to human-centric AI development.

Conclusion

Recital sixty-one of the EU AI Act represents a critical acknowledgment of the potential risks associated with the deployment of AI in alternative dispute resolution settings. By classifying these technologies as high-risk, the EU seeks to establish a regulatory framework that prioritizes human rights, accountability, and transparency. As AI continues to permeate various sectors, the implications of this classification will shape not only the future of ADR but also broader discussions surrounding the ethical use of AI in legal systems.

Therefore, in the European territory it will not be possible to conduct, solve legal disputes and even less to render decisions, including those concerning alternative dispute resolution procedures, by AI itself or alone. The use of AI as tools in the legal sector is, and in fact, shall be permissible, to enhance and contribute to the legal sector, but never to exclude human performance from the equation.

As humans, we were created with unique “Humics.”  Each Domain Name Case requires the application of such Humics on a case-by-case basis, and even more, if the legal dispute is related to mediation, arbitration, or a particular legal technical opinion. Therefore, Real Human Supervision and Balance must be the approach in terms of the use of AI in the legal sector.

About the author

María Alejandra LÓPEZ GARCÍA is a bi-lingual Intellectual Property Attorney, with extensive experience in Domain Name Dispute Resolution, Trademark Monitoring, IP enforcement, including Social Media, Marketplaces, Paid Searches, and Mobile Apps. She is a UDRP Panelist and qualified to practice law in Venezuela since 1999; and she has an LLM in Industrial, Intellectual Property and Information Society Law from the University of Alicante, Spain.

YouTube Channel: iprsonline

___________________________________________________________

[1] WIPO Guide to the Uniform Domain Name Dispute Resolution Policy (UDRP), https://www.wipo.int/amc/en/domains/guide/#a1

[2] WIPO Internet Domain Name Process, https://www.wipo.int/amc/en/processes/process1/

[3] Terminator I – 1984, or Terminator II, Judgment Day – 1991.

[4] BORNET, Pascal, What AI will never replicate: The Humics, IRREPLACEABLE with AI (April 16, 2024), https://pascalbornet.substack.com/p/what-ai-will-never-replicate-the

[5] BORNET, Pascal, Combating Skill Obsolescence: The Humics Approach, IRREPLACEABLE with AI (Oct. 9, 2024), https://www.linkedin.com/pulse/combating-skill-obsolescence-humics-approach-pascal-bornet-uotee/

[6] Ibid.

[7] Pride (In the Name of Love), U2, 1984.

[8] UUK, Risto et al., EU Artificial Intelligence Act (Regulation (EU) 2024/1689), (Nov. 1, 2024), https://artificialintelligenceact.eu/annex/3/

[9] UUK, Risto et al., EU Artificial Intelligence Act (Regulation (EU) 2024/1689), (Nov. 2, 2024), https://artificialintelligenceact.eu/recital/61/

[10] WHITEHEAD, Ronald, Explainable AI: Understanding the Black Box (Oct. 26, 2023), Institute of Analytics, https://ioaglobal.org/blog/explainable-ai-understanding-the-black-box/#