Rowan Insights
Focus on Rowan AI Functionality – February 2024

Last month, we talked about what our clients have been telling us about their Gen AI/LLM plans. This month, we will provide a look at what we’ve developed to date.

Before we get into our plans, the USPTO recently released guidance relating to the use of AI in patent applications. This memo recognizes the potential benefits of the use of AI, but also makes it clear that AI cannot reduce or replace the ethical responsibilities of the drafting attorney.

At Rowan, we think about how our integrated patent drafting environment can assist patent practitioners with improving the quality of their work product, and the efficiency of their workflows. We think of AI as being extremely helpful to help them generate and transform content that they can shape and mold into their work product, not to completely replace the process of creating it.

From our discussions with our clients, Rowan adopted three imperatives for our AI/LLM work:

1. Ensure it is based on our client input. As with everything we do at Rowan, we turn to our clients to guide our efforts. We like to build technology not based on an abstract idea, but on actual feedback from our clients. In our development process, we build new functionality and give our clients the opportunity to test drive it in a beta version before we release it publicly.
2. Make it optional. We have made our AI functionality configurable. It can be turned on or off at the organization level, as well as at the user level. So that those that are ready can begin utilizing it, but those that want to hold off can do so.
3. Be transparent. We will not send information to a remote LLM service without the explicit instruction of the user.

Last fall, Rowan added the “Rowan AI Chat” functionality to our Rowan Products integrated patent drafting environment. Activated optionally for those clients who choose it, this window assists the practitioner in querying ChatGPT through their API to provide content that can be inserted into the patent application.

Integration through the API has several benefits. Queries using the API as implemented by Rowan are NOT used to train OpenAI’s LLMs as the terms of service from OpenAI assure confidentiality of data. Alternatively, queries in the web interface to ChatGPT may be observed by OpenAI employees and may be used to train future models.

Additionally, while using the OpenAI web interface to ChatGPT, the user must switch windows and copy/paste content from the application to the chat window and then back to the application. By using Rowan AI Chat instead, the practitioner can stay in the same application and use automated controls to share information between the chat and the drafting environment.

Last fall, Rowan added the “Rowan AI Chat” functionality to our Rowan Products integrated patent drafting environment.

As an example of how Rowan AI Chat works, if the patent application is related to 6G cellular service, the practitioner could ask the chat for background on 6G cellular technology. The practitioner can then review the content provided, and if satisfied, insert it into the application with one click. At this point, the practitioner can use the advanced drafting capabilities of Rowan Patents to revise and transform the content into a work product that is ready to be filed. If the practitioner feels the response needs another try, there is a button to regenerate the input based on considering additional content.

In this fashion, Rowan Patents allows the practitioner to reap the benefits of ChatGPT while still remaining responsible for their final work product, in line with the USPTO guidance about the ethical responsibility for the drafting attorney.

As of today, Rowan is releasing a further enhancement to the Rowan AI Chat — Prompt Assistance. Prompt Assistance automatically searches the draft application for content related to the prompt. That related content is added to the prompt, providing more context and improving the resulting generated text. As always, the user is in full control of the text being sent to the Large Language Model.

Next month, we’ll look ahead to our future plans and where we are going with these exciting technology advancements.

If you would like to receive data on applications filed by your firm or your company like the below, to help you understand the power of analytical data to help you, please contact us and we will be happy to analyze your applications as a courtesy.