The use of generative artificial intelligence (AI) is quickly becoming nearly universal, and it has profoundly changed how law firms approach advanced problem-solving and strategic decision-making. From drafting internal memos to summarizing dense legal documents, these tools deliver an undeniable boost to efficiency.
However, for the CEO or general counsel, this convenience presents a significant, largely uncharted legal hazard. The central question facing leadership today is whether using these platforms may accidentally destroy the very protections that keep your company’s legal strategy confidential.
Attorneys must focus on assisting clients in navigating the intersection of law and innovation. CEOs and other stakeholders must understand that a business strategy is only as strong as its ability to remain protected from competitors and adversaries.
As courts begin to weigh in on how attorney-client privilege and AI interactions should be treated (not to mention other areas, like AI and copyright infringement), the results have been anything but uniform. Interpreting these shifting legal sands is now a requirement for any executive who wants to preserve a competitive advantage.
The Core Conflict: Tool or Third Party?
The threshold issue for any court is how to characterize the AI itself. If an AI platform is viewed simply as a technological tool, like a word processor or a cloud storage service, then using it might not waive any protections. However, if the court views the AI as a third party capable of receiving a disclosure, the results can be catastrophic for your case.
In one noteworthy criminal case, a federal court in New York ruled that exchanges between a defendant and a popular AI platform weren’t protected. The court argued that because the system is not an attorney (despite its having a reasonable understanding of legal principles), no privileged relationship could exist.
Additionally, because the defendant freely shared information with a third-party platform, the court found that he had no reasonable expectation of confidentiality. This decision serves as a clear warning: Legal AI confidentiality risks are not just theoretical. They are real and, crucially, can lead to the forced disclosure of a legal team’s sensitive defense strategies.
Conversely, a federal court in Michigan reached a different conclusion in a civil matter. That court characterized generative AI programs as tools rather than persons. It noted that if the mere act of using a modern drafting environment waived protection, it would nullify work product rules for almost every modern office. While this provides some comfort, it also highlights the lack of a national consensus on the question, “Does AI waive attorney-client privilege?”
Privacy Policies and the Death of Confidentiality
One of the most dangerous traps for a business is the fine print in an AI platform’s terms of service. Most consumer-facing AI systems reserve the right to collect user inputs to train their models. Some even state that they may disclose user data to third parties, including government authorities, in certain circumstances.
When a court examines legal issues and AI confidentiality risks, it often looks at whether the user took reasonable steps to keep the information private. If your employees are using free, public versions of AI platforms, they may be consenting to data sharing that effectively invites others into your private strategy sessions.
Importantly, courts have noted that when a platform’s policy clearly states that it collects and may share data, a user can’t reasonably claim they expected the conversation to remain secret. At this point, technology users do (or at least should) know better. This is why the question of whether AI waives the attorney-client privilege often turns on the specific technical settings and contracts you have in place.
The Critical Significance of Counsel Oversight
Industry experts and ethical guidelines now highlight the need for a "human in the loop" when using these technologies. This is not just a matter of guaranteeing the AI doesn't hallucinate facts, but also an essential part of preserving legal protections. Recent rulings suggest that if a client uses an AI tool on their own volition without the direction of an attorney, the chances of maintaining privilege are slim.
Recent ABA guidance for AI lawyers highlights the duty of confidentiality and the requirement of technological competence. This means that if you want to use AI to summarize legal research or generate ideas used to formulate a strategy, it is much safer to do so under the explicit direction of your legal team.
When an attorney directs the use of a tool, it may fall under the work-product doctrine or be treated as the lawyer's agent, though this remains a developing area of the law. Ignoring the ABA guidance regarding AI could leave your firm vulnerable to discovery requests that you had not anticipated.
AI in Litigation Discovery: Protecting Your Strategy
The danger of AI misuse often comes to light during the discovery phase of litigation. Opposing legal teams are now specifically looking for signs that a party used AI to prepare their case. If an attorney identifies a summary or memorandum created by an AI, it can open the door to more intense scrutiny. This can result in broader requests for all AI prompts and outputs. Consequently, it is crucial to understand the implications of AI in the litigation discovery process.
Additionally, e-discovery AI risks include the possibility that confidential data was uploaded to a cloud platform without proper encryption or contractual safeguards. If a court finds that the use of a tool was not relevant, you might be protected. However, the safer route is to minimize your company’s AI-related litigation discovery exposure through strict usage policies and secure, enterprise-grade systems.
Safeguarding Your Intellectual Assets
To reduce e-discovery AI risks and protect your company’s privilege, leadership must take forward-looking measures.
- Implement detailed AI usage policies. You should clearly define which AI tools are approved for business use and strictly prohibit the entry of any sensitive or classified information into public systems.
- Invest in enterprise-grade AI solutions. These platforms frequently include contractual guarantees regarding data privacy and ensure that your inputs are not used to train the public model.
- Implement mandatory employee training. Your team must understand that a chatbot might be considered a third party and that sharing strategy with it can be as damaging as accidentally blind copying someone on a sensitive email.
- Keep your legal team involved in all AI initiatives. Ensuring that AI is used at counsel's direction can provide an additional layer of protection under the work-product doctrine.
Our Approach to AI and Litigation
At Martensen, we believe that technology should be a competitive weapon, not a liability. We work with clients to develop intellectual property strategies that account for these emerging risks. Whether you are dealing with IP in government contracts or defense supply chains, or your focus is high-stakes litigation, our attorneys have the technical and legal experience to ensure your innovations remain protected.
Managing the intersection of attorney-client privilege, AI, and corporate strategy requires more than just general knowledge of the latest release of your platform of choice. It requires a partner who understands how to build a strong IP portfolio while steering clear of the potential pitfalls that new technology can create.
If you are concerned about how your team is using AI or want to ensure your current litigation strategy is secure, contact us today.