ByteSize: The growing use of Artificial Intelligence dispute resolution

Current applications and potential use of AI in international arbitration

In the realm of current applications, AI systems are revolutionising data analytics and document review by processing vast data sets to identify trends and patterns. This technology is also facilitating document production, where it swiftly identifies relevant documents, redacts sensitive information, and sorts out privileged or confidential documents that should not be disclosed. AI’s role extends to document analysis, offering insights into case-relevant information, and real-time interpretation services, which translate spoken language instantaneously, thus overcoming language barriers.

Another notable application is the generation of factual summaries, where AI condenses large volumes of data, making complex information more accessible to practitioners.
Looking ahead, AI holds the promise of predicting judicial outcomes, thus assisting arbitration professionals in formulating strategies based on historical data. It could also streamline the selection and appointment of arbitrators by identifying individuals with the necessary expertise and background. Furthermore, AI might play a supportive role in decision-making processes by providing data-driven insights.

Limitations, challenges and risks

These advancements are not without challenges, in particular in relation to:

Legal Reasoning, Accuracy and Reliability: AI systems lack the capability for legal reasoning, leading to errors, inaccuracies and potentially oversimplified outcomes. In addition, AI may generate non-factual information - ‘hallucinations’ - which could mislead practitioners and potentially affect the integrity of legal proceedings. Outdated data is also likely to produce inaccurate results, and AI's potential inconsistency in providing the same answer to repeated queries raises questions about its reliability

Training and Testing:  The multiplicity of laws and diverse practice areas as well as the confidentiality of most commercial arbitration awards make it challenging to train and test AI systems effectively for arbitration;

Confidentiality and Data Privacy:  The handling of sensitive information by AI raises concerns about maintaining confidentiality and protecting data privacy, especially when dealing with privileged information

Ethical Implications: The use of AI necessitates careful consideration of ethical issues, such as bias and fairness in decision-making processes. In particular, the data used to train AI models can reflect human biases. If not addressed, these biases can be perpetuated and amplified by AI, leading to skewed and unfair responses and decision-making processes

Equality of Arms: Access to advanced AI tools could create imbalances in legal proceedings and infringe on the principle of equality of arms between the parties

Transparency and ‘Explainability’: AI systems do not inherently provide the reasoning behind their decisions, which is a fundamental aspect of legal adjudication. In high-stakes legal disputes, the reasoning behind decisions is crucial. It adds legitimacy, helps parties understand the outcome, and guides future behaviour and legal rulings. Without transparent reasoning, AI’s role in legal decision-making is therefore limited. In addition, legal decisions serve as precedents that influence future cases. AI’s current inability to articulate reasons for its decisions hinders its capacity to contribute to the development of legal principles and precedents

Validity and Enforcement of AI-Influenced Awards: The legal acceptance of AI-influenced decisions is uncertain, especially if the technology is restricted in certain jurisdictions

AI Initiatives in judicial fields

In response to these issues, several initiatives have been undertaken or are under development. The European Ethical Charter on the Use of Artificial Intelligence in Judicial Systems and their Environment (2018) and the Supreme People’s Court’s Opinions on Regulating and Strengthening the Applications of Artificial Intelligence in the Judicial Fields (2022), for instance, offer framework and guidance for the application of AI in judicial fields.

In August 2023, the Silicon Valley Arbitration and Mediation Center, a forum specialising in technology-related dispute resolution, released its draft Guidelines on Artificial Intelligence in Arbitration for public consultation. 

Commentary

In the field of litigation, we are seeing similar benefits from the use of AI. Software has been tested to help with decision-making. FraUKe (Frankfurter Urteils-Konfigurator Elektronisch), for instance, has been tested by the Regional Court of Frankfurt and is intended to assist in decision-making for mass proceedings of air passenger claims.

From a UK perspective, Judicial Guidance was produced towards the end of 2023 by a cross-jurisdictional judicial group to assist the judiciary, their clerks and other support staff on the use of AI.  There is hope that AI will enable a faster and more cost effective digital justice system going forward.  It will also be interesting to see how the House of Lords’ Communications and Digital Committee Report recommendations (including support for copyright holders and exploring options for an ‘in-house’ sovereign UK large language model) and the outcome of the Information Commissioner’s Office consultation on generative AI models will shape the landscape of AI in the dispute resolution context going forward.

Looking forward

The use of AI in dispute resolution will only continue to expand, with new tools and approaches, and ever greater levels of sophistication. 

As for AI as the subject matter of the dispute, there are several points for organisations to consider in the procurement, development and deployment of AI – please see our recent AI guide.

Update on the Product Security and Telecommunications Infrastructure Act 2022 and accompanying 2023 Regulations

Update

Following our December update, a reminder that the Product Security and Telecommunications Infrastructure Act 2022 and the accompanying 2023 Regulations comes into force on 29 April 2024. 

The over-arching purpose of the legislation is to “make consumer connectable products (“smart” products) more secure against cyber-attacks”.

Manufacturers, Importers and Distributors of UK consumer connectable products will have duties including to: (i) not supply products into the UK that do not meet the security requirements contained in the Regulations; and (ii) respond appropriately when notified that a product supplied by them is not compliant with the Regulations.

A product may meet the definition of a UK consumer connectable product even if it is solely aimed at business customers, e.g. if identical products have been made available to UK consumers by another distributor.

Manufacturers must ensure UK consumer connectable products they supply comply with the security requirements at Schedule 1 of the Regulations, which: 

1. prohibit the use of universal default passwords and ‘easily guessable’ passwords
2. require Manufacturers to establish a “clear, accessible and transparent” system for security issues with UK consumer connectable products to be reported; and
3. require Manufacturers to publish the defined support period for each UK consumer connectable product capable of receiving security updates

A ‘Manufacturer’ includes anyone who has the product manufactured for them or markets a product produced by a third party under their own branding. Importers and Distributors must not supply a UK consumer connectable product into the UK without a statement of compliance prepared by the Manufacturer certifying that the product is compliant with the Regulations, or a summary of the statement of compliance.

For a fuller summary of the new legislation, please see our December update, and please get in touch with any specific questions.