English 
Español 
Artificial Intelligence (AI) should be viewed as another technological tool that lawyers can use in the provision of their services, similar to what the typewriter, the telex, and the fax were, or others like the email, and legal research applications available on the Internet. When using it, they must comply with their ethical obligations, including protecting their clients’ confidential information.
We begin by noting that the latest iteration of the Panamanian Code of Ethics was adopted in November 2022 (the previous one was approved in January 2011), almost at the same time as the launch of ChatGPT in its first version.
Every lawyer is under the obligation to “employ in [the attention of a business] all lawful means, seeking to act with rectitude, honesty, discretion, and respect.” (See Article 6) Likewise, the lawyer is “obliged to use all his or her knowledge” (Article 12) and “must obtain a thorough understanding of his or her client’s case before advising them on it.” (Article 8). Finally, Article 26 is categorical: “The lawyer’s responsibility is of a personal and direct nature…”
The aforementioned Panamanian regulations require lawyers to use the lawful resources and means available. Not doing so could, in itself, be a violation of their ethical obligations. Therefore, they are responsible for doing so to the extent that AI allows them to lawyer more effectively in favor of their client’s cause.
However, they must do so with rigor and honesty, which implies, for example, giving credit to the AI tool, just as they would when quoting a work by Dulio Arroyo or Arturo Hoyos (the image accompanying this article was created with Copilot). They are also responsible for reviewing the AI product prior to its use, just as they would when dealing with the product of a colleague or subordinate, whether before a counterparty or before an authority or judge. Legal professionals, by training, must use their judgment when obtaining information provided by any tool, including AI tools.
In the section “On the practice of the profession” in the Preamble to the Code of Ethics and Professional Responsibility of the Lawyer of the Republic of Panama, it is noted that the lawyer’s obligation to “respect the inviolability of the information and documents entrusted to him.” Article 13 of the same Code of Ethics imposes on the lawyer the duty to “keep the secrets and confidences of their client,” and paragraph (7) of Article 37 of the Code prescribes that it is unethical for anyone to “disclose, violating professional secrecy, any confidence made to them by their client or third parties, unless authorized by the latter or the former or done with a direct aim at their own defense.” Therefore, the lawyer must ensure to have the minimum competencies in the use of AI, both in terms of what one contributes and in relation to the product.
Most publicly available AI tools (ChatGPT, Gemini, etc.) expressly warn that they do not guarantee the confidentiality of the information you provide (input via prompt). On the contrary, these tools are fed and ‘built’ on what they ‘know’ from publicly available information and what the user provides. The legal professional must know if the AI tool they will use guarantees that the information shared by their client will not become public domain. In short, before using AI, each lawyer should verify that the tool he or she uses allows him or her to fulfill ethical obligations of confidentiality. It is true that the use of publicly accessible AI tools can pose a risk when it comes to confidentiality and data protection. Once this information is provided to systems like Chat GPT, the system may reuse that information, but not necessarily with the person who provided it initially, but with anyone who uses the system in general. It should be emphasized that this risk can be mitigated by using AI tools configured for private access (for example, by integrating the limited use of Copilot). In short, it is due to this latent risk that transparency with the client regarding the use of this tool must prevail, especially if it is publicly accessible, and diligence in the use of the tool, since the tool should make the work of humans more efficient, not replace it. Unless authorized by their client, the lawyer would be violating the Code of Ethics if he or she provides confidential client information to those public tools.
We must also not lose sight of general concepts, such as the one established in Article 5 of the Lawyer’s Code of Ethics, which states that the lawyer “must act with honesty and good faith. He must not advise, tolerate, or resort to fraudulent acts; nor affirm or deny falsely.” It is well known that the output of these tools is not 100% accurate or correct, and that they have even ‘hallucinated’ (less so than in their initial versions).
We do not find any limitations in the Code of Ethics regarding the use of AI by lawyers, subject to the previously mentioned warnings and with respect to the prohibition of plagiarizing their colleagues (art. 37(31)) and the obligation of procedural good faith. AI is one of those resources, making it clear that the product of AI must be subject to review and analysis by the lawyer before its use, whether with the client, with the opposing party, or with any authority or judge. In fact, since AI (at least at this moment) cannot distinguish what is “truth,” the lawyer should not use information obtained from AI without verification; if they do, they could potentially violate Article 37(33) of the Code of Ethics in case their ‘allegations are false,’ among other things.
It should also be noted that a lawyer’s legal fees are also subject to regulation of the practice of law. The Code of Ethics clearly states that a lawyer may only charge “reasonable compensation” (Article 11). According to Article 15, among the factors that guide the determination of professional fees are “the work required and the nature of the case” and “the degree of participation of the lawyer in the study, presentation, and development of the matter.”
A particular challenge will arise when fees are agreed upon based on the time invested, since with AI, lawyers will be able to produce the same work in less time; do they have to pass that savings on to the client? Can they charge a fee for the use of AI? In our view, it is appropriate to do so to avoid infringing ethical standards. It has always been an ethical breach to ‘invent’ time to justify fees or deliberately delay the completion of a task.
There are already AI applications that draft or review contracts, refine or formulate arguments, and even generate legal opinions, as well as assist in jurisprudence research, but there is even talk of the possibility of predicting how a judge will rule based on available information about them and previous rulings. In short, just as happened with some predecessor technologies, lawyers must be very aware of the ethical implications of using AI.
In Colombia, a family judge already used AI as part of their process to support their ruling, which was challenged for violating due process. The Supreme Court of that country declared that the way the judge relied on the AI tool did not constitute a violation of due process.
In our country, on October 11 of this year, the new Civil Procedure Code (approved by Law No. 402 on October 9, 2023) will come into effect, which explicitly provides in Article 172 for the use of technology to facilitate and expedite access to justice.
In this regulation, it is established that procedural acts may be carried out using AI, but its use cannot replace the human reasoning of the judge or magistrate. Thus, it is stated: this is a tool and not a substitute for human resources. While, for the moment, there may be uncertainty regarding how its use will materialize in procedural acts, there should be no doubt that the explicit recognition of these tools in a regulation is an indication that AI is influencing the practice of law in Panama and, of course, it will be one of the present challenges faced by the administration of justice and the auxiliaries of the judicial system, in addition to those other branches of law where the use of AI is in vogue.