Rise of the machines: will lawyers be replaced by Artificial Intelligence? What do the services providers have to offer?

Lígia Padovani

Lefosse, São Paulo

ligia.padovani@lefosse.com

Conference report

Biennial IBA Latin American Regional Forum Conference: Technology, social media and artificial intelligence: challenges for the legal industry in the digital age

23 March 2023

Plenary session

Session Co-Chairs

Adriana Castro BLP Abogados, San Jose; Special Projects Offi cer, IBA Latin American Regional Forum

Hans Sydow SLG Consultores, Caracas; Publication and Newsletter Offi cer, IBA Latin American Regional Forum

Speakers

Dario Cadena Lloreda Camacho, Bogota

Heath Harris EY Law, New York City, NY

Andres Jara Alster, Santiago

Andrew M Levine Debevoise & Plimpton, New York City, NY

Rosa Rascon Thomson Reuters, Mexico City

This session discussed current developments in artificial intelligence (AI) and its use, impacts and concerns in the legal sector. The two co-chairs and four panellists began the discussion with an overview of the advances of AI from a legal perspective and the main related issues, chaired by Hans Sydow. This was followed by technical topics commented on by the speakers.

Overview: advances in AI from a legal perspective

The main issues considered by the session were those involving the use of AI tools within legal practice and business. According to Hans Sydow, AI is being faced and discussed from a more practical approach, that is, as a source of rational analysis for data processing. The understanding of such an approach leads to new and expanded discussions on the implications of AI, seen in recent debates over certain tools, such as ChatGPT.

According to Sydow, from a legal perspective, the main concerns relating to the application of AI are the need for a better understanding of how legal practitioners can use such machines and tools in their work, and how to train lawyers properly to use them to their advantage.

For him, the training of machines centres on two main issues that need to be addressed and solved. These are: the possibility of emergent behaviour (ie, behaviour that has not been purposefully programmed to the tool, but emerges from a consequence of the system), causing the AI tools to become unpredictable; and the difficulty of confirming if the facts and information on which the AI tool based its response are in fact true and correct.

Relevant topics relating to the application of AI in the legal sector

The second part of the session focused on the following topics which are relevant to the subject.

The use of AI in the legal sector

The first topic was introduced by Andrew M Levine, who recognised AI as the next chapter, and that the ability of machines to create content is extremely helpful.

Levine commented on the popularity of ChatGPT and its many possible uses, including through the preparation of press releases. He also talked about AI having already been used for years in some legal areas, such as predictive coding for litigation matters.

Other examples of AI tools that are being used in the legal frame were also mentioned. These include: the use of platforms by companies to process data received and sent by email; the analysis of all content and knowledge already gathered over time, and the avoidance of unnecessary engagement by legal firms.

Heath Harris stressed the advantages of having machines condensing and processing data, such as for the analysis of high quantities of contracts and for comparison of different markets and potential approaches. Such tasks would take days or months if undertaken by humans but can now be performed in minutes. The speed of development and capability of platforms demonstrate the power of AI to transform how legal work is performed and how legal practitioners create value for clients.

According to Levine, there are exciting new platforms under development. In compliance areas, AI can be a useful tool for processing data and identifying, almost instantly, potential frauds, corruption, and the transaction’s main red flags in respect to anti-corruption.

Despite the advantages of using AI in the legal field, panellists also described AI as a source of tremendous error and problems, which was further discussed in the following topics.

The challenges of AI

The possibility of machines providing incorrect or false information was commented on by most panellists. In line with the challenges described in the overview, it was affirmed that answers provided by AI are only as good as the information on which they are based.

Rosa Rascon stated that it is important for everyone to acknowledge that AI is not magic and that there is a need to train the machines, a task which must be performed well. Similarly, Andres Jara discussed the need to manage knowledge in order to provide information to AI tools correctly, and fill in the gaps that machines need to function.

Other than the intrinsic challenges of AI, new legal concerns have also arisen from its development, specifically concerning intellectual property created by machines and the ability to protect the generated content. Although current legislation around the world grants copyrights only to humans, it is believed that the evolution of AI will also stimulate the granting of copyright to machines and that there will be similar debates in future.

The impact of AI on the legal profession

The speakers also discussed the impact of AI on the legal profession, emphasising the distinction and possible redefinition of legal practice and legal business.

For Harris, experienced lawyers will continue to be necessary at the ‘top’ of the knowledge pyramid, to interpret processed data and effectively practice law. However, at ‘mid-level’ and the knowledge connected with and arising from it can be redefined by AI, hopefully democratising knowledge and enabling the presence of different individuals, with different educational backgrounds, to participate in legal business. In similar terms, Rascon highlighted the ability of AI to create new positions, such as legal engineers and data scientists.

The possibility of using AI not only for legal practice, but for administrative tasks and matters relating to the legal environment was also commented on. Outsourced legal services providers seem to be dominating this initiative in the legal field, by harnessing technology and AI for performing administrative tasks in a cost-effective manner.

In this scenario, lawyers and legal practitioners should focus on understanding, at least in a high-level basis, how AI tools work and which tools are being used by service providers, as to identify and adapt them to the reality of their life and work, and to innovate within their law firms or legal departments.

Finally, the general message was not to be afraid, but to harness, step-by-step, the use of AI for the benefit of legal practitioners.

Ethical implications and liability

The last topic discussed in the session was the ethical implications and liability issues arising from the use of AI in the legal field.

Andrew M Levine identified four ethical principles that are closely connected to the application of AI: confidence; communication; confidentiality; and supervision.

The confidence principle was described as the obligation of individuals to take charge of their actions and to provide the best service possible to clients. As tools develop, it is important to legal professionals to know the benefits they may bring and to implement them effectively to improve rendered services.

In respect to communication, the need to inform and discuss with the client the use of any AI tools was highlighted, including their related risks.

The confidentially obligation was indicated as one of the major concerns with respect to ethics, considering the risks of leaks and the lack of protection of certain platforms. On this matter, the need to implement policies relating to ChatGPT and AI in general was stressed. This should at least comprise restrictions on uploading confidential information and trade secrets to any AI tools.

On supervision, panellists discussed the need of legal practitioners to oversee the application of technology and AI in services provided to clients; and, to the extent possible, diligence to make sure that the platforms used do not breach any legal obligation and are not biased, especially if applied to human resources and the recruitment of personnel.

Lastly, commenting on potential liability arising from the application of AI tools, Dario Cadena recognised the challenges and uncertainties concerning the matter, and stressed that it is essential to bear in mind that there are responsibilities and ethical guidelines which need to be complied with. To summarise, the application of AI must take place under the principles of not harming, being fair and operating transparently.