How AI Tech Suppliers Can Address IP Lawyers' Concerns

• Safeguarding Confidentiality of Information

Confidentiality is essential to law practice, especially when it comes to IP. An IP lawyer may know about the most cutting-edge inventions and concepts months or years before they are ready for public announcement.

Title 37 of the U.S. Code of Federal Regulations, Section 11.106, covers confidentiality of information, requiring lawyers to make reasonable efforts to prevent inadvertent or unauthorized disclosure. Under this provision, the lawyer cannot reveal anything related to the representation of the client without obtaining the client's informed consent. Also, they must make reasonable efforts to prevent unintentional disclosure.

The AI tech provider should ensure that privilege, data privacy, and confidentiality are preserved when their AI tool is being used.

To address the confidentiality concern, the AI tech supplier can ensure that separate infrastructure, including servers and hard drives, are used to shield confidential data from unauthorized parties. This must include security measures preventing external entities from accessing the data. It can also encompass ethical walls that block access to data within a particular law firm or corporate entity that has two cases that need to be kept separate. Cross-sharing of data should be prohibited unless permission is granted by the IP lawyer or client.

• Preventing Unauthorized Access to Information

When using an AI tool on a specific matter, the lawyer is typically exposing the client's confidential material to the AI supplier in some way.

To gain authorization for sharing data with AI tool providers, IP lawyers need to inform the client explicitly—in writing—that they are using a particular AI tool. This can often be done up-front in the client engagement letter. When soliciting the client's written approval, the lawyer must disclose all necessary information so that the client can knowingly approve the disclosure of information to the AI tech provider.

The client should have the option to opt out of AI tech use, too.

The AI tech supplier contributes to this process by providing clear, explanatory language in plain English about the AI technology, indicating the scope and nature of their tools' data usage. If a client expresses concern, the AI provider can simply restrict use of the AI product to using only publicly available data, such as U.S. Patent and Trademark Office (USPTO) filings and office actions.

• Restricting Use of Publicly Available AI Tools

Eventually, apps such as ChatGPT, Gemini, and Perplexity will likely become just as ubiquitous as the Google search bar to find information.

Lawyers are becoming more comfortable using AI search engines, and they may be running documents through tools like ChatGPT to see what outcomes it produces.

However, IP lawyers using these platforms without understanding their ramifications are taking a risk. Use of publicly available tools like ChatGPT should be completely restricted to only publicly available information. Lawyers are advised against experimenting with free AI tools because they don't know where their—and their clients'—information is being stored and used.

Accordingly, lawyers are advised to use only the subscription versions of AI tools for legal work, because at least that gives some sense that information entered there will be kept private. Free versions are certainly not private, so there is no expectation of confidentiality there.

AI tool suppliers have a responsibility to clearly explain to consumers where the data is traveling to during use and if it is being stored.

• Providing Closed-Loop Systems

For lawyers who want to use AI tools, the best-case scenario is a closed-loop or sandboxed system, which is a completely separate server hosted by a known tech services provider.

Also, the AI tool must not be programmed to learn from any data provided to it. This is a limitation for sure, but this way, the cases are shielded from each other, and the AI cannot cross-learn.

Ultimately, IP lawyers need to feel safe about using the AI and be confident that their information is not going anywhere else. This extra element of precaution allows IP lawyers to use AI services for more in-depth purposes, such as drafting patent descriptions and other legal documents.

• Preventing Exposure of Trade Secrets

IP practitioners are responsible for conducting appropriate fact-gathering from their clients, which includes understanding trade secret protection. As such, IP lawyers must apply specific protective measures to safeguard trade secrets.

Trade secrets may be packed into various inventions and their disclosures, but they may not appear in patent applications. Trade secrets should never touch the AI tools at all. It's up to the lawyer to discern which parts of an invention or an invention disclosure are trade secrets that should be totally excluded from exposure to the AI tools.

The AI tech supplier can cooperate by adding warnings to lawyers to prevent them from uploading materials containing trade secret details—but the burden is ultimately on the lawyer to ask the correct questions and understand the operation of trade secret laws.

Another caveat is the distinction between patent agents—who can only practice before the USPTO—and registered patent attorneys who are also members of a state bar association. Patent agents aren't licensed attorneys and can't give legal advice concerning trade secrets, but this does not mean they shouldn't be well-informed to spot potential trade secret issues.

• Establishing Competency Using AI Tools

If lawyers are going to use AI tools, they must understand how to use them properly. This speaks to competency under Title 37 of the Code of Federal Regulations, Section 11.101.

The USPTO's concept of competency lies in lawyers' backgrounds but also in their ability to use tools properly when representing clients. Before using an AI tool on client work, the lawyer must consider whether they are competent to use that tool and whether they have received training on it.

The news headlines about lawyers submitting briefs full of AI hallucinations are cautionary tales that show the importance of technical competency.

IP lawyers must play a supervisory role, understanding the limits of AI tools and pursuing adequate and reasonable training to use them properly. This includes not only their own usage of AI but also that of their junior attorneys—senior lawyers need to be able to spot AI-generated content. AI tech providers can offer online video tutorials, 24/7 chatbots and knowledge bases, tech support, and consultants who can provide necessary training for judicious use of their solutions

• Ensuring the AI Provider is Vetted and Qualified

AI tech suppliers should be carefully selected to ensure compliance with relevant data security and related standards. The AI company's IT team should explicitly attest to the performance of their products against strict IT security, data protection, and safeguarding policies.

Vendors can distinguish themselves by establishing, implementing, and maintaining International Organization for Standardization (ISO) security standards. Typically, tech suppliers who have passed these highly recognized evaluations will indicate these distinctions on their website and in sales proposals.

AI tools are becoming a permanent fixture in today's business world, so IP lawyers should leverage them to be more productive and effective. However, lawyers must also ensure that their use of AI is consistent with their legal ethics responsibilities.

By collaborating closely with vetted AI tech providers, IP lawyers can maximize the upside of AI use and reap the benefits, while simultaneously upholding their professional accountability.

To find out more about the ethics of AI in patent practice, watch our on-demand webinar 'The Ethical & Practical Use of AI in Patent Prosecution,' featuring Registered Patent Attorney Kevin Bronson, one of the authors of this article.

This article was first published on Law360 (subscription required) on November 21, 2025.