New York Law Journal October 6, 2025
By Jonathan Bick Jonathan Bick is counsel at Brach Eichler in Roseland, and chairman of the firm’s patent, intellectual property, and information technology group. He is also an adjunct professor at Pace and Rutgers law school
Attorneys are increasingly aware of the benefits of using artificial intelligence (AI) in their law practice. However, attorney’s awareness of AI inherent effects on ethical obligations is lagging. Applying ethics rules learned from the integration of the Internet in the practice of law will ameliorate AI ethical risks difficulties.
AI related attorney assistance products are abundant and used in most law practices. According to a recent Bloomberg Law survey of in-house and law firm attorneys 63% of all survey respondents reported that they have used AI in some way for work.
However, few attorneys understand the technology underlying AI and even fewer still recognize the potential risks that its ever-present use in their practice poses to their ethical obligations. A brief survey of recent news articles related AI associated breaches of the rules for professional responsibility found attorneys bad acts included: using AI as a substitute rather than an assistant; failing to verify all AI outputs before relying on them professionally; using AI to engage in the unauthorized practice of law by proxy; and failing to preserve confidentiality due to neglecting protected client information while using third-party AI platforms.
Fortunately, the adoption of AI technologies is similar to the Internet explosion twenty-five years ago. Just as attorneys then learned that they must apply ethics rules to the new Internet tool in their practices, and attorneys using AI today must do the same. Both then and now attorneys have an ethical duty to make sure that every associated attorney, paralegal, or other person who could be doing work with AI appreciates what AI can and cannot do, what problems can transpire, and what to do when an AI issue does arise.
As in the ethical adoption of the Internet into law practice, the first step in ethical adoption of AI into law practice is understanding the nature of the technology. In the case of the Internet, lawyers needed to understand that the Internet was protocol governed (i.e. by software technology) rather than by a central governing authority. From a technology perspective, attorneys must understand that generative AI is software which generates content in response to prompts, wherein said content that is statistically most likely to be acceptable is sent as a reply to a prompt.
An AI response is not based on the generative AI’s search of the Internet for the answer to the prompt, but rather on the data upon which the AI software was trained. The data upon which the AI was trained is significantly limited by the data, computational resources, costs, and ethical considerations related to the AI machine learning element used.
In short, AI machine learning acknowledges common-sense understanding, so they can only make predictions based solely on statistical patterns without genuine comprehension. While machine learning models can identify correlations and patterns in data, they are not capable of responses associated with true understanding. Thus, AI is susceptible to making predictions based on spurious correlations.
Just as attorneys learned that the Internet was a set of communicating computers sharing a common protocol, attorneys must learn that AI is computer software which is confirmed to respond to a given set of human-defined objectives, to generate content in an automated manner; and use model inference to formulate options for information.
Due to the limited data and training AI systems receive, AI are capable of a multitude of error which may result in ethical violations. The more significant errors from an ethical perspective include “hallucinations,” or inaccurate or incomplete output, and exposing confidential firm or client information to unauthorized persons. These errors may expose the firm or its lawyers to lawsuits alleging malpractice, breach of fiduciary duty, or other claims, and related violations of ethical rules.
While using the Internet, attorneys learn about the ethics of using the Internet from a mix of primary materials, such as state ethics codes and ethics opinions, and articles about legal ethics and professionalism, especially short articles about legal ethics and the Internet, and news articles from on-line legal journals. The same sources may be used to learn about AI ethical use.
Since both the Internet and AI are computer and software driven, they share many of the same ethical issues. In both instances, a client’s consent is needed to use either the Internet or an AI, for example, where confidentiality is a concern. In both instances, existing ethical opinions state that merely adding general boiler-plate provisions to engagement letters is not sufficient to meet this requirement.
Additionally, both Internet use and AI use may result in other violations of disciplinary rules of professional conduct. For example, AI may result in a lawyer accepting employment in a legal matter which the lawyer knows or should know is beyond the lawyer’s competence. The ABA has issued opinions related to the use of the Internet and the use of AI that prompts lawyers that they are required to have a reasonable understanding of the capabilities, limitations, benefits, and risks of using the Internet and AI, respectively.
Procedures implemented to ensure the application ethical duties for Internet use are applicable for AI use. For example, since both the use of the Internet and AI in law practices may require input of client information, the procedures adopted by lawyers for safeguarding information relating to Internet communication during the representation of a client is adoptable to ensuring that AI systems handling confidential data adhere to confidentiality obligations.
The same operating procedures related to conflict of interest and candor toward tribunals as implemented for the use of the Internet may be applied to the use of AI. During both Internet use and AI use lawyers must identify and address potential conflicts of interest arising from using the technology. While using the Internet or AI, attorneys must ensure the accuracy and relevance of the citations provided by the technology that is used in legal documents or arguments.
The same measures used by an attorney supervising those using the Internet to ensure compliance with disciplinary rules must be applied to AI use. A lawyer is subject to discipline because of another lawyer’s violation of the rules of professional conduct, if the lawyer is a supervising lawyer and encourages or knowingly permits the impermissible conduct involved.
In other words, because of the risk of inaccurate output by the Internet and so-called “hallucinations,” an AI independent verification of the output is necessary. Users of the Internet and AI must evaluate the technology being used, analyze the output, not rely solely on the technology’s conclusions, and understand that technology cannot replace the lawyer’s judgment.
Such supervision for both Internet and AI use applies to the unauthorized practice of law. Attorneys using the Internet or AI must ensure that they are not assisting someone in the unauthorized practice of law. To the extent an AI programmer is not a lawyer and is providing legal analyses, that programmer may violate ABA Rule 5 regarding the unauthorized practice of law.
At least sixteen states or courts have issued reports, ethics opinions, or judicial decisions focused on the ethical considerations inherent in the use of AI. Each of these communications appear to have an Internet communication covering the same issues but issued decades ago.