introduction
Artificial intelligence (“AI”) tools are rapidly becoming pervasive in the modern workplace, fundamentally changing the way lawyers and other professionals perform their jobs and communicate with each other. AI-enhanced note-taking applications are now commonplace in corporate offices and law firms alike. Although these technologies may offer efficiencies, many organizations adopt them without fully understanding the significant legal risks they pose.
Perhaps the most pressing concern for attorneys and their clients is the potential violation of attorney-client privilege. When AI tools process, store, and transmit privileged attorney-client communications (especially when disclosed to third-party cloud providers), they can inadvertently waive the confidentiality protections that protect those communications from discovery. Apart from communications with lawyers, clients’ use of AI note-taking tools creates a large volume of potentially discoverable business records, exposes candid conversations that were previously unrecorded, and increases the complexity of discovery due to the sheer volume of data. The use of AI note-taking applications also involves the lawyer’s ethical considerations to ensure that the use of AI tools, including note-taking applications, is consistent with core ethical obligations such as duties of competency, independent judgment, loyalty, and confidentiality.
Recent state bar opinions and court decisions have highlighted the risks associated with using AI note-taking tools in legal settings. As this technology continues to evolve and become more integrated into daily practice, attorneys must remain vigilant to understand how these tools work, where client data travels, what safeguards exist to protect privileged and confidential information, and how to advise clients regarding the use of such tools.
State attorney’s opinion on ethical issues regarding recording conversations with clients
On December 22, 2025, the New York City Bar Association issued a formal opinion (the “Opinion”) regarding the ethical obligations of attorneys under the New York Rules of Professional Conduct when using AI note-taking applications. This opinion believes that ethical issues are fundamental to a lawyer’s professional responsibility, regardless of jurisdiction.
A lawyer’s duty of loyalty requires him or her not to engage in any act that involves fraud, fraud, deception, or misrepresentation. This has been interpreted by the New York City Bar Association to mean that attorneys are required to obtain the client’s consent to use recording devices, even in states where only one-party consent is required. Clients expect confidentiality when speaking with their lawyers and may choose their words more carefully knowing they are being recorded. However, if the lawyer does not disclose it, the client is deprived of the option of disclosing it.
Lawyers must also adhere to confidentiality obligations when using AI note-taking programs and retaining the information they generate. Lawyers should consider the privacy and security safeguards put in place by tools, including what data is stored, how long it is retained, rights to deletion, and whether such tools are used to train AI models. To meet ethical obligations, it is important to advise clients of the risks of losing confidentiality and attorney-client privilege protections.
The New York City Bar Association also emphasizes that attorneys’ competency obligations prohibit them from relying solely on AI artifacts. All minutes and summaries should be reviewed for accuracy, and lawyers should not rely solely on AI artifacts. As a general matter, lawyers should understand the technical features, limitations, and security of the tools they and their clients use, and be aware of the ethical issues that arise from such use. Training and supervision of subordinate attorneys and other employees should include similar practices to ensure that practices consistent with attorneys’ ethical obligations are applied universally throughout the firm.
In this regard, the opinion acknowledges the problems inherent in the representation of clients by lawyers using proprietary AI tools. When clients use these programs independently, attorneys have little control over recordings and summaries. To mitigate negative consequences, the opinion recommends that attorneys should address the possibility of such use at the beginning of the relationship. The New York City Bar Association recommends that attorneys require that conversations between attorneys and clients not be recorded when AI tools are used with their clients, include clauses in business agreements that AI work product is not dispositional or binding unless reviewed for accuracy by the firm, and provide clear and consistent disclosures about the risks of loss of confidentiality and privilege.
Separately, lawyers must consider not only the ethical obligations owed to their clients, but also legal and regulatory compliance regarding the use of these AI tools. Recording a meeting without the consent of all participants may violate wiretapping laws. Consent requirements vary, but some states require the consent of all participants before recording a meeting. California, Connecticut, Florida, Illinois, Maryland, Massachusetts, Nevada, New Hampshire, and Pennsylvania are among these bilateral agreement states. AI note-taking applications may not have the appropriate notification settings required in two-party consent situations where all parties must consent to recording. Additionally, some states have laws that protect biometric data, including “voiceprints” derived from the audio files used by these tools to generate transcripts and summaries. AI tools alone cannot reliably determine the legality of recordings in all jurisdictions, so lawyers must be aware of that to ensure compliance.
Otter.AI Privacy Lawsuit
Otter.AI Privacy Litigation, No. 5:25VC06911 (ND Cal., August 15, 2025) illustrates the potential confidentiality risks when lawyers and clients alike use note-taking applications. The case stems from a California federal court case against Otter.AI, an AI note-taker application that the plaintiffs allege recorded and transcribed conversations without proper consent from all participants in a meeting.
The Otter.AI Privacy Policy places the onus on users to obtain consent. Additionally, the privacy policy at issue states that Otter.AI may share your personal information with third parties, including “cloud service providers…data labeling service providers…providers of integrated third-party programs, apps, or platforms,” and that users may “ [its] service, [users] allow [it] To share specified information and data created and/or uploaded [users] to [its] We provide servers with these third-party programs on your behalf. ” The use of AI note-taking and transcription tools with similar policies could defeat attorney-client privilege protections and raise ethical issues for breach of confidentiality.
America vs. Heppner
America vs. HeppnerNo. 25-cr-00503-JSR (SDNY, October 28, 2025), highlights the concerns raised in the New York City Bar Association opinion paper when clients use AI in legal matters without the direction or control of an attorney. heppner We addressed whether documents created by criminal defendants using AI platforms are protected by attorney-client privilege and work product. Defendants charged with securities fraud and related charges independently used Claude, Anthropic’s generative AI tool, without the guidance of their attorneys, to create a report outlining potential litigation defense strategies, which they later shared with their attorneys.
The court rejected the privilege and work product claims, raising an issue similar to the New York City Bar Association’s concern that the documents were not confidential, given that Anthropic’s privacy policy explicitly authorized data collection and disclosure to third parties, including government authorities. This decision has significant implications for the use of AI platforms with similar disclosure policies, suggesting that entering privileged information into such systems could erode attorney-client privilege and work product, effectively waiving confidentiality protections.
However, heppner Although the court’s decision concerned third-party tools whose public policy showed insufficient protection of output confidentiality, the court did not determine whether the use of corporate or in-house AI tools under attorney direction with appropriate confidentiality safeguards for such work falls within the scope of attorney-client privilege. Therefore, it is important to consider whether an AI note-taking tool provides sufficient protections to be considered a privileged attorney-client communication tool.
Practical considerations
Lawyers should approach AI note-taking applications with intentional care and comprehensive safeguards.
- Obtain client consent for lawyers’ use of AI tools: Lawyers should inform clients that an AI note-taker or transcription tool may be used during the engagement, what precautions have been taken to protect confidentiality, and address any concerns expressed by the client prior to use. In practice, this notice is often included in the engagement letter. Additionally, some of these tools can provide notification and request consent at the beginning of each meeting where the AI tool is used. We recommend that you use such notifications if they are available.
- Check the output of all AI tools. All notes and summaries must be reviewed for accuracy by your attorney. Lawyers cannot provide a competent defense if they rely solely on the output of an AI tool without review.
- Make your clients aware of the risks of using AI note-taking tools alone. Lawyers should advise clients that the use of third-party AI note-taking tools alone may create records that are not protected by privilege or work product doctrine, where such records may be discovered in litigation. Advise clients not to use AI note-taking tools without the guidance of their attorney in any capacity to discuss ongoing litigation matters, including legal strategy and computation of damages.
- Establish clear internal policies and screening procedures. Businesses and customers should develop policies governing the use of AI notetakers, including validation steps to ensure accuracy. Organizations should carefully vet their AI tools to ensure they implement closed systems, do not have third-party data distribution, and do not use client data to train the underlying models. In-house lawyers should consider internal policies that limit the type of information that can be recorded by AI note-taking apps and prohibit recording discussions about legal or sensitive confidential matters.
- Cross-functional governance: A team of technology, compliance, and legal experts should review and approve AI tools with recording capabilities to enhance safety measures.
- Prepare for discoverability: Lawyers should advise clients on retention procedures for records and summaries produced by AI note-taking tools, including what data can be retained and for how long, and the development of procedures to ensure that legal holds extend to such information, where appropriate.
