We are sharing the full interview with Andrés Chirgwin on AI in the legal profession: “The most important limit concerns replacing decision-making”
Andrés Chirgwin, managing partner at Chirgwin, spoke with Actualidad Jurídica, DOE’s blog, about the challenges that artificial intelligence poses for the legal profession. During the interview, he discussed the cognitive and educational risks associated with the unmindful use of these tools, the boundaries between technological support and professional responsibility, the safeguards that legal teams should adopt, and the skills that could make a difference in an increasingly automated legal market.
The adoption of artificial intelligence, automation, and digital tools is changing how lawyers and law firms work. Generally speaking, what are the main challenges these technologies currently pose for the legal profession?
I would say there are several types of challenges, but perhaps the most fundamental one is a medium-term challenge: the cognitive and emotional risks posed by artificial intelligence.
This is also related to the structure of law firms and the education of new generations. These issues are connected, although they are not entirely the same.
The unmindful use of artificial intelligence can lead to cognitive and educational deterioration. Ultimately, human beings tend toward rest, and the brain is an organ that uses a great deal of energy. Therefore, there is a natural tendency to overuse tools that can replace aspects that are important to a person’s cognitive development and the exercise of their mental abilities.
In that sense, the legal industry faces the risk that lawyers may become worse lawyers. I would say that this poses a risk to the market and to institutions.
Moreover, many of the functions traditionally performed by junior lawyers, such as research or preparing initial drafts, are among those most susceptible to replacement by artificial intelligence. This may affect both employment opportunities for new generations and their education.
In the medium term, this could lead to a deterioration in the overall standard of the profession, both because of the unmindful use of artificial intelligence and because of a reduction in the number of well-trained lawyers.
In my view, those are the major challenges. There are other, more practical, day-to-day challenges, but these are the main ones.
One of the current debates concerns how to benefit from the efficiency offered by these tools without replacing lawyers’ legal judgment. Where do you believe the line should be drawn between technological support and professional responsibility?
Professional responsibility will always remain.
When legal services are provided or legal reports are prepared within an organization, there is still someone who signs off on them. That signatory remains responsible, because an important part of their value lies precisely in signing off: in saying, “I validate this” or “I affirm this.”
That remains true whether the work was completed using artificial intelligence, without artificial intelligence, or with some assistance from artificial intelligence.
The risk arising from the misuse of AI, or from its use without verification, is that it increases the likelihood of errors. At the same time, lacking the skills to use these tools may also diminish a lawyer’s ability to add value, regardless of their area of practice.
The most important limit concerns replacing decision-making. That is the key point.
One recent example that has been discussed involved a court decision in which the prompt that had been used was leaked. Part of that prompt instructed the artificial intelligence to decide the case. That is precisely what must never be done, under any circumstances.
When that happens, practically all the work has already been delegated. The human element is removed from the decision, which, in the context of administering justice, is philosophically very complex.
Using artificial intelligence as a support tool is one thing; disengaging from the intellectual exercise and allowing the machine to do everything is quite another. That is where the line must be drawn.
These new technologies also create risks concerning personal data protection and information security. What safeguards should legal teams adopt before incorporating digital platforms or AI tools into their daily practice?
I would say there are two issues.
The first is personal data protection, particularly now, with the new law. It is a stringent law based on the European model that establishes obligations and risks for companies and for all types of persons that process data.
Law firms, like any other company, are required to comply with those rules.
In the particular case of law firms, however, client confidentiality and the duty of professional secrecy also come into play. This obligation is specific to certain professions, including the legal profession, and requires more serious and stringent measures concerning the systems used and the manner in which they are used.
Many artificial intelligence tools learn from the questions or information they receive, creating risks associated with data being retrieved in response to new queries.
This can result in negligent, rather than intentional, breaches of a lawyer’s duty of professional secrecy.
For this reason, I would recommend that law firms and legal departments conduct a thorough technical review of the tools they use, their obligations, and how those tools relate to regulatory compliance and professional secrecy.
At a minimum, the systems must undergo a technical and legal review before being incorporated into daily work.
Looking ahead, which skills do you believe will be most important for lawyers in an environment shaped by artificial intelligence?
The legal industry is one of the industries most susceptible, in many respects, to being replaced by artificial intelligence tools.
It is often said that more strategic and multifaceted skills will be more difficult to automate. I am referring to the abilities that make it possible to integrate disparate information in order to provide clients with better advice.
Complex matters, intangible analyses, and anything involving strategic understanding will be more difficult to automate than purely technical legal work, such as drafting a contract or a court filing.
There may also be opportunities for lawyers with empathy or emotion-recognition skills, particularly in terms of perceiving what machines are still unable to fully grasp.
What will always remain, however, is the hallmark of accountability. Increasingly, I believe clients will ask for someone to be accountable for an opinion or advice, so that a company or an individual can make decisions on that basis.
Systems do not do that. You cannot sue ChatGPT for negligence arising from legal advice.
Therefore, professional responsibility will remain. What will change considerably is how that opinion is reached.
