Scope Of AI In ADR Decision Making And Developing Laws
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
SCOPE OF AI IN ADR DECISION MAKING AND
DEVELOPING LAWS
Surya Sundaresan, Assistant Professor, St. Joseph’s College of Law, Bangalore Sameeksha TR, St. Joseph’s College of Law, Bangalore Scope of AI In ADR Decision Making and Developing Laws We are moving towards Artificial Intelligence into everything, this Artificial Intelligence will soon give us an artificial world to live in, looking back to the past we had technology replacing man where we use to have manual interference for everything that we have, work did not seem to be work it was a part of life gradually with every step towards development we became more and more dependent on the technology. Now how technology can be put on a negative front over AI or artificial intelligence is that the amount of manual efforts is still not completely washed off as AI to an extent is still dependent on man as an entity to survive. The World around us changes in seconds and with every change we need to change ourselves but what this evolving world is moving to would not provide fairness and justice and this article extends this concept further to how inclusion of AI can be threat to the legal system overall. As when we narrow our discussion down to the legal field, we have different branches where the inclusion of AI is acceptable and can also be said that it is the need of the hour but this article constricts itself to the inclusion of AI into dispute resolution and it demanding some amount of human intervention which leads to a question on justice, fairness, and bias. It covers a detailed idea about how negative it can be if we include AI into dispute resolution as even the feeding is done with human involvement it gives a very high scope to biases discission based on pre fed data and this article also addresses the question that is it practically possible to have pre fed decision from these AI tools that are currently in use in the near future generation when we count on precedents as our major source for dispute resolution cases. .
INTRODUCTION
Evolving into a world full of AI is not only virtuous but equally be a challenge to bias and privacy. Whatever has been alleged as a myth from the experience humans have in the past it is still an experiment whether we can include AI with zero human intervention and expect accurate result in all the sectors of its inclusion it is true that it has been successful in a lot of spheres such as where it can be an assistant for timely completion and faster results but overtaking humans and zero intervention is still a challenge that has to be answered. The major apprehension that we have developed is that the world where we now live have a voluntary acceptance of AI using our data be it personal or professional the gap of something private is no more in existence. Now this gathering of information could be on approval and a lot of times would be a just recording of certain gestures or actions which collects data for analysis later. Now this is where the actual problem starts as we understand what our mind things would be certain times different from what our body would have been acting and our thoughts would have been on a different track but this explanation or understanding becomes unnoticed when this recording of our day-to-day actions happens and this downside is important to be understood and to be rectified. The fast-paced development of Artificial Intelligence has reshaped our lives, becoming integral to industries and activities that once required significant human participation but the drawbacks is something that we tend to forget. It could be because of widespread adoption of AI systems that it has sparked increased discussions about fairness and bias in Artificial intelligence, as biases and discrimination become more noticeable. Now that we discuss about pendency and delivery of justice and the delay in the same we need to work on these drawbacks and include this magical mechanism of AI in decision making and inclusion of these systems with the other developing laws and regulation as it will open a plethora of ideas to monitor the preparation and to take cautious steps, to check and calculate the action and outcome, to have a precalculated documentation in certain commercial actions and the most important inclusion in ADR mechanism. It has been a challenge for all of us in the legal and alternative dispute resolution (ADR)
fields to pilot the hype and potential of AI. The use of AI in ADR has the potential to improve efficiency, reduce costs, and enhance the overall effectiveness of the dispute resolution system. However, this needs a certain level of human intervention to address complex issues of justice, fairness, and bias. As these AI systems increasingly influence outcomes in ADR mechanisms, questions arise regarding their ability to uphold equitable standards and prevent biases. As a lot of times the data collected by these AI systems as discussed before would be voluntary actions from day-to-day surveillance or it could be an outcome of unintentional actions that have been on record and has been a seed to decision making by the AI. The understanding that these are not legal setbacks but have to be improved through scientific means is the challenge. How can a proper understanding as to what forms a part of the source for decision making and this segregation to be coded and decoded by the system itself is the achievement that we can look towards in future. Now when we talk of decision making it is very important that we discuss about inclusion of AI in the legal system be it in administration or delivery of justice and how it can be a threat to the entire system how it can even be a help in future, how this human AI duo that we have can a lot of times lead to biased decisions and lack of accuracy can result in wastage of more time and efforts.
AI AND LEGAL SYSTEM
When we start with AI in legal system the ground is vast to discuss each specialization of law. As the paper focuses on decision making and the bias in that regard we restrict ourself to the dispute resolution concept as according to the authors this is the segment which is developing and can be applied to any field of law be it criminal or civil as in the coming times we look forward to inclusion of ADR mechanism in criminal laws so resulting to this understanding the use of AI in dispute resolution has gained considerable attention and prominence in recent years be it from the apex courts or lower courts wherein we see suggestions of adopting such mechanism. AI technology offers immense potential for enhancing efficiency and accuracy in resolving disputes. From chatbots that can quickly answer basic legal queries to sophisticated algorithms that analyse massive amounts of data and recognise patterns, AI has the potential to revolutionize how disputes are managed and resolved. This could be with the help of pre fed
data not by human intervention but by everyday surveillance that we are a part of in our daily life and just utilization of this data to analyse the best possible settlement in certain cases and to come up with the best possible negotiation term for both the parties. While AI provides significant advantages in streamlining dispute resolution processes, it is crucial to recognize its limitations and potential downsides. One of the primary concerns is the issue of bias. AI systems are trained on historical data, which may contain inherent biases. If these biases are not identified and addressed, AI tools could perpetuate or even exacerbate existing inequalities. This is particularly problematic in dispute resolution, where fairness and impartiality are paramount. The complexity and sensitivity of many disputes require a deep understanding of context, emotion, and ethical considerations that AI may not adequately capture. Therefore, while AI brings remarkable benefits to dispute resolution, it also introduces potential negative effects such as perpetuating bias and compromising the intricacies of human judgement. This is where as suggested before inclusion of surveillance methods and data usage history will play a positive role in negating all this bias claims to some extent at least.
INCLUSION OF AI CAN BE A THREAT TO THE LEGAL SYSTEM
The integration of artificial intelligence into legal technology offers numerous benefits but also presents ethical challenges. AI can enhance efficiency, but it lacks human like intuition. This raises concerns about its ability to understand complex legal issues and human behaviour. One such major concern is the risk of perpetuating biases inherent in the data used to train AI algorithms. If these biases left unchecked, it will lead to unjust outcomes and it will lead to inequalities within the justice system. The use of AI in legal proceedings, like replacing judges with AI, raises fundamental questions about fairness1, equity2 and accountability3. It will be difficult for us to understand how decisions were made, potentially undermining the principles of due process. Along with this, AI system may struggle to uphold human rights and ensure equal treatment for all individuals. 1 Barocas, S., & Selbst, A.D. (2016). Big Data’s Disparate Impact. California Law Review, 104(3), 671-732.
2 Eaglin, J. M. (2017). Constructing Recidivism Risk. Emory Law Journal, 67(1), 59-122.
3 Citron, D.K. (2008). Technological Due Process, Washington University Law Review, 85(6), 1249-1313.
AI involves several threats when integrated into legal technology. One significant concern is the potential for AI to figure out the biases present in the data it learns from. If the training data used to develop AI algorithms contains biases, these biases can be amplified in the AI’s decision- making processes, leading to unjust outcomes and reinforcing existing in equalities with the justice system4. AI technologies can enhance efficiency and accuracy, their integration demands a certain level of human intervention to ensure fairness and justice. Depending on human oversight raises questions about the extent to which decisions made by AI align with legal and ethical standards. One major concern is the potential for AI systems to perpetuate biases present in the data they are trained on. If AI algorithms learn from historical data that reflects societal biases or systematic inequalities, they may inadvertently produce biased outcomes in dispute resolution. Moreover, the use of AI in dispute resolution may lead to a loss of trust in fairness of outcomes. The individuals involved in disputes may question the impartiality5 of AI decisions and the extent to which human judgement and empathy are taken into account. Therefore, AI offers potential benefits in streamlining dispute resolution, careful consideration must be given to how it is implemented to ensure that justice and fairness are upheld.
HUMAN PRE-FED DATA LEADS TO BIASED DECISIONS IN DISPUTE
RESOLUTIONS.
A bias is “a tendency, inclination, or prejudice towards or against something or
someone”6. Since AI is created by humans, it is not free from bias. Programmers may unintentionally or intentionally embed their personal biases into algorithms. Consequently, some level of bias is often observed in these algorithms, leading to unfair outcomes. Bias can infiltrate AI systems in various ways, particularly through their algorithms. AI system learn from training data, which may contain biased human decisions or reflect historical and social inequalities. Additionally, the bias of the human programmers who design these systems can be embedded in the algorithms, consciously or unconsciously. Lack of representation also contributes to bias. If certain groups are not represented in the programming process, the AI will fail to recognize their experiences, resulting in a system 4 Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019), Dissecting racial bias in an algorithm used to manage the health of populations, 447-453.
5 Eaglin, J. M. (2017), Constructing Recidivism Risk, Emory Law Journal, 59-122.
6 Psychology, 2021
biased towards a limited segment of the population and lacking inclusivity. It is not that an AI system completely free of bias cannot be developed. It is theoretically possible to build a system that makes impartial decisions based on data. However, the quality of these decisions is only as good as the quality of the input data. Thus, achieving complete impartiality in AI systems is not expected soon. Humans tend to show preference for certain individuals, groups, or things, often developing a special affection or support for them. It’s common for people to have a favourite teacher, friend, or family member. AI systems are trained on historical data fed to these systems, which often reflects societal and institutional biases. This can lead to AI reproducing and even amplifying these biases in decision-making processes, negatively affecting the fairness and impartiality expected in dispute resolution. When the training data contains biases related to race, gender, socioeconomic status, or other demographic factors, the AI system is likely to perpetuate these biases. The negative impacts of biased AI decisions in dispute resolution are profound. They can result in unfair outcomes, diminish public trust in the legal system, and lead to increased legal challenges. Once the biases are embedded in AI’s algorithms, they can be challenging to detect and correct. This is because AI systems learn from patterns in the data, even without rigorous oversight and continuous updating of the training datasets. Biased outcomes can persist and even worsen over time. One of the significant negative effects of incorporating AI into dispute resolution is the absence of “Speaking orders”, which are reasoned orders explaining the rationale behind judgements. AI-generated decisions typically lack this essential reasoning, leaving litigants without a clear understanding of how a decision was made. This issue arises because AI systems are pre-programmed by statisticians and experts who assign weight to various pieces of evidence based on past trends, potentially skewing results depending on the training data sample used. The lack of transparency in AI judgements further exacerbates the problem by making it difficult to build trust in AI-powered dispute resolution. Since, AI’s impartiality depends heavily on the data it is trained on, dissatisfied litigants often hesitate to use AI for future cases, not knowing why their case was unsuccessful. This opacity can deter early adopters and slow down the integration of AI into the legal industry.
Additionally, without a clear demonstration that AI processes are based on fair and impartial data, and without improved transparency in AI decision-making, the adoption of AI in dispute resolution remains problematic. Although AI has the potential to resolve simple, low-value disputes, these significant drawbacks suggest that its integration into the legal industry may be as beneficial as anticipated. The negative effects of lacking reasoned orders and transparency underscore the challenges that must be addressed before AI can be effectively implemented in dispute resolution.
CONCLUSION
AI has significant potential as a future means of dispute resolution is the final concluding note that the authors would like to start their conclusion with now this is because ADR mechanism is itself at a stage of hit and trial in the Indian Legal system now what could be a plus to this is that if we include AI in this initial stage itself we can rectify the discussed bias and also incorporate the developments that can be recognised in the experiment stage. While currently in its nascent stage, AI is improving daily and holds much promise. Just as virtual courts have become the new normal, AI could similarly be integrated into the legal system with a few key developments to enhance transparency and eliminate bias. These improvements could ensure that small disputes are resolved quickly, without making litigants feel deprived of justice. The legal system is already burdened with numerous cases, and AI could provide an alternative means of resolution. To address the current shortcomings of AI in dispute resolution, several suggestions can be considered to transform its weaknesses into strengths. Firstly, by advancing research to enhance lie detection and eliminate data bias, AI’s reliability and fairness can be significantly improved. Secondly, increasing transparency in AI decision-making would boost litigants’ confidence, enabling quicker, cheaper, and more consistent dispute resolutions. Further, integrating AI with minimal human intervention during the early stages can help guide the AI to ensure general decisions are made correctly. This approach, though seemingly counterintuitive, could ease AI’s introduction into the legal system by providing a human presence, thereby increasing comfort and trust among users. While AI is not yet the optimal tool for dispute resolution with time and adequate
research, its main drawbacks can be resolved. AI has immense potential in dispute resolution, particularly in cases where a clear decision is needed. Improved transparency and minimized bias will make it easier for litigants to trust AI, paving the way for its broader adoption in the future, even if not in mediation or sectors focused on amicable settlements. AI will be most beneficial in areas requiring definitive judgements, establishing itself as a valuable tool in the legal field. One example that we can conclude with is that especially in commercial disputes and resolutions the outcome for settlement is based on data and profit so here inclusion of AI is one example to such progress just a small thought over the other possibilities could be a path to no human intervention in decision making hence no chances of potential bias.
REFERENCES
1. Rishab Sundar, Scope of AI in Alternative Dispute Resolution, International Journal of
Law Management & Humanities, Vol. 5 Iss 5, pg. 1034-1037.
2. Solhchi, M.A., & Baghbanno, F. (2023). Artificial Intelligence and its Role in the
Development of Future of Arbitration, International Journal of Law in Changing World,
2 (2), 56-57,
3. David L. Evans, Joshua Walker, Principles Supporting the Use of AI in Alternative
Dispute Resolution, International Centre for dispute Resolution, 2023 AAAi lab principles press release.indd (adr.org)
4. Chaudhary, G., Artificial Intelligence: Copyright and Authorship/ Ownership
Dilemma? (2022) Indian Journal of Law and Justice, pp. 212-238.
5. Dr. Priyanka Jawalkar, Alternate Dispute Resolution and Artificial Intelligence,
International Journal of Creative Research Thoughts (IJCRT), Vol. 10, Iss 12, pg. 2230- 2882, 2022.