Artificial intelligence: Promising or (Un)Promising?

By Devangi Dube
2021 Essay Scholarship 1st Runner Up
 

The topic for the essay reads, “Whether the introduction of artificial intelligence in the legal profession will mitigate bias, increase access to justice, and promote diversity in the society.”, and I strongly believe that it will. Before delving into the arguments, it is instrumental to break down the topic for a more well-rounded understanding of the same.

According to the dictionary, “Artificial Intelligence” is the development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. AI in the legal profession enhances an attorney’s ability to research, advise, and serve their clients. AI systems are constructing learning models to predict the outcomes of cases, using as inputs the corpus of relevant precedent and a case's particular fact pattern. This essay would essentially analyse specific implications of the introduction of AI in the justice system. These revolve around answering 3 main questions – Firstly, how long does it go in “mitigating bias” or alleviating the pervasiveness of prejudice in the dispensation of justice; Secondly, does it increase “access to justice” or the number of people that can gain from services the legal occupation has to offer; And finally, does it play a role in “promoting diversity” and nurturing the heterogeneity in society?

I realised that the answers to these questions are not binary or absolute. Therefore, it is important that I clarify that the stance I undertook emerged from a spectral analysis with due acknowledgement to the benefits and drawbacks alike. The topic should be interpreted not as a mere advantages and disadvantages question but rather as whether the advantages outweigh the disadvantages.

The introduction of AI can assist in mitigating bias, thereby encouraging greater inclusivity and promoting diversity in society. More often than not, it is challenging to come up with a job posting which accurately represents the organisation and job requirements, while also attracting a not only qualified but also diverse candidate pool. In this filed, the potential of AI can be unlocked. AI-enabled job posting review technology can detect bias in draft advertisements before they appear in front of potential candidates. The AI can highlight gender, age and ethnicity biased language, while also guiding on tone, voice to enable re-wording for a posting that will appeal to a wider range of qualified candidates. Furthermore, during the recruitment process, hiring officers seldom know everything there is to know about the candidates and rely on cues in their applications which signal to personal attributes, which leads to greater propensity of such officers producing skewed judgements. In this case, AI that is trained with appropriate algorithms can make recommendations to either remove, or replace with neutral terms, any wording that may lead to biased judgments. This particular usage of AI can be applied specifically to the Legal profession as well. Thus, by mitigating bias in and diversifying the reserve of legal professionals who play a definitive role in redefining social roles, one can diversify the society at large as well.

The opponents of the proposed suggestions might argue that AI in itself does not possess a “sense” of objective fairness since it acts on and learns from algorithm that is created by humans, thus reflecting human biases by default. However, what opponents fail to acknowledge is that since AI does not just do tasks it has been programmed for and actually learns as it goes, improving performance through feedback, such biases can be overcome. The same can be achieved if we enable and empower a culture that expects workers to prioritise equity at every step of algorithm development. Keeping in mind that total elimination of bias may not be feasible, organisations should uphold standards of accountability around the working of their models. The inherent AI bias can also be overcome if we have diverse teams researching, developing and operationalising algorithms. These diverse teams can include individuals belonging to different gender, race, religion, age, and also legal experts that understand the legal domain the AI is meant to operate in. It is also important to understand that bias in AI is not merely a technical issue but also arises from societal inequities that can be addressed by responsible leaders in the legal profession who understand this larger, holistic view and the “fairness” that can be at play. Such leaders can use their influence to support industry reforms. Moreover, the public must have access to the manner in which AI is created and improved to better comprehend the results of the data processes that impact their lives, enabling them to contest decisions made by algorithms. Another solution is to look at sectors where bias is a much smaller factor. The key in choosing the right legal area is focusing on areas with strong rules as opposed to standards. The former provides the ability to have clear feature engineering while that latter requires specificity to train an accurate model. As a proponent of introduction of AI in the legal profession, while acknowledging the drawbacks associated with it, I attempt to show the possibility that the same can be dealt with in an effective manner. Thus, proving that the unfathomable potential of AI in the legal sector far surpasses the challenges it brings with it.

“Our intelligence is what makes us human, and AI is an extension of that quality”
-Yann LeCun, NYU

While we resonate with the aforementioned quote, let us understand that increased access to justice includes not only an increase in the number of people who can access the services, but also an increase in the quality of those services. AI provides cost effective resolutions by identifying the legal infirmities in judgements, assisting in drafting contracts, due diligence, legal analytics, etc. In this way, AI can act as a catalyst in alleviating the burden of the judiciary, especially in those cases to be decided by human judges. Thus, judges can focus their brain power where it is needed the most, increasing the quality of the work produced by them. We realise that the essence of AI is not to supplant the human brain or the judges but to facilitate them to reassess the processes they follow, the work they do and to ensure their outcomes are more predictable and consistent, ultimately providing wider and access to justice to the common citizens. 

To conclude, I would like to say that law is an avenue that represents a significant opportunity for value creation. Emerging technologies must be a primary focus of the next generation of attorneys as they will unlock innumerable possibilities to transform and revitalise the field of law. Legal professionals must address their new responsibilities, support and promote ethical principles to keep pace with the changing times and positively impact the relevant regulatory frameworks. Automation won’t be replacing lawyers any time soon, but as AI evolves it will be able to increasingly assist the people in the industry, to become more educated about their options and use their time more efficiently. At the same time, AI needs a strong legal framework to explore its maximum benefits. As I remain optimistic about its prospects, it is the balance between AI and the people who regulate it that will be interesting to watch.