Generative Artificial Intelligence (AI), such as ChatGPT, is changing the way legal professionals perform their daily tasks, from drafting documents to conducting legal research. However, this technological advancement carries significant risks that lawyers and other legal professionals must take into account when using these tools. In this article, we explore the dangers of blindly relying on generative AI, illustrated by a court case in Spain, and offer recommendations for its safe use in the legal field.
Key Risks of Using Generative AI in the Legal Sector
1. False or Erroneous Court Summonses
One of the most obvious risks of using generative AI in the legal field is the possibility of obtaining false or incorrect legal citations. AI, even when advanced, can generate erroneous information that appears authentic at first glance—a phenomenon known as “AIhallucinations.”It is essential that lawyers verify all sources and citations generated by AI before incorporating them into legal documents.
2. Inappropriate foreign case law
Generative AI tools may provide references to foreign regulations that are not applicable under national law. This can result in the inclusion of irrelevant legal citations in local proceedings, which could compromise the validity of the documents and affect the outcome of the case. It is crucial that legal professionals use AI with caution and adapt the results to local regulations and case law.
3. Leakage of Confidential Data
The use of generative AI on open platforms poses significant risks to privacy and the confidentiality of information. By providing sensitive data to these platforms, there is a risk that it could be used to train future AI models, which could lead to the leakage of confidential information. Legal professionals must exercise extreme caution when entering sensitive data into these tools and consider the privacy implications.
4. Overreliance on AI
Although generative AI can automate legal tasks, such as drafting documents or searching for case law, it should not replace professional oversight. While 44% of legal tasks can be automated using AI, not all of them are suitable for being completely delegated to automated systems. Critical decisions must remain the responsibility of lawyers, who must apply their knowledge and professional judgment to ensure the quality and accuracy of legal work.
An Example of the Dangers of Generative AI in the Legal Profession
Recently, a Spanish lawyer filed a complaint that included a legal citation from the Colombian Penal Code. As he himself acknowledged in a subsequent letter, the reference was provided by ChatGPT3, and he committed this “gross and unintentional factual error” “due to his firm’s improper use of the artificial intelligence system, ”for which he apologized in writing.
The Supreme Court of Justice opened a separate proceeding to investigate whether the attorney had committed an abuse of rights or acted in bad faith during the proceedings. Despite the error, the court decided to dismiss the case without imposing financial penalties, noting the attorney’s prompt apology and the novelty of the issue.
The Supreme Court of Justice opened a separate proceeding to investigate whether the attorney had committed an abuse of rights or acted in bad faith during the proceedings. Despite the error, the court decided to dismiss the case without imposing financial penalties, noting the attorney’s prompt apology and the novelty of the issue.
Generative AI: A Revolution, but Not a Replacement for Legal Professionals
This ruling underscores the importance of verifying all information provided by generative AI tools and issues an ethical warning to legal professionals: “May this serve—perhaps—as a warning regarding the legal, professional, and ethical implications that the careless use of new technologies may entail.”
Generative AI is a revolution because of its processing power, but under no circumstances should it replace professionals, since the ability to filter information and adapt it to a specific legal context remains a task exclusive to lawyers.
Recommendations for the Safe Use of Generative AI in the Legal Field
- Verification of Sources. It is essential to always confirm that the information provided by AI is accurate, relevant, and up-to-date. It is recommended to use only reliable sources and to verify that any legal citations are applicable to local law.
- Data Protection. Confidential or sensitive information should not be included on generative AI platforms, especially those that are publicly accessible. Additionally, it is recommended to use only AI tools that guarantee high standards of privacy and security.
- Professional Oversight. Although AI can facilitate many tasks, it should not replace professional judgment. Attorneys must review and approve all work generated by these tools before using it in their practice.
- Continuing education. Legal professionals must stay up to date on best practices and advancements in AI, as well as the legal and ethical implications of its use in the legal field.
Generative artificial intelligence offers many opportunities to improve efficiency in the legal sector, but it also poses significant risks that should not be ignored. Cases such as the recent incident in Spain serve as a warning to all legal professionals about the dangers of blindly relying on these tools.
At ELZABURU, we are committed to the responsible use of technology in the legal sector, always under the supervision and guidance of our professionals. AI can be a great ally, but the final decision must always rest with the legal expert.