Mastering Generative AI: A Lawyer’s Guide

Last week, Ozan Yalti, DraftWise Co-founder, and Tony Mauriello, General Counsel, joined Theresa Spartichino, Sr. Manager of Practice Technology at Ropes & Gray, for a conversation hosted by the American Bar Association, on Mastering Generative AI: A Lawyer’s Guide. 

Generative AI has transformed the legal landscape, and transactional lawyers are increasingly faced with the challenge of understanding, evaluating, and effectively using these new tools and skills in their practice. In this webinar, the DraftWise team discussed how lawyers can approach generative AI with a strategic mindset—focusing on education, evaluating emerging technologies, and upskilling to stay ahead of the curve. 

Here are the insights from the conversation: 

What is Gen AI?

The last few years have brought a significant leap in artificial intelligence with the transition from traditional AI to Generative AI (Gen AI). Before Gen AI, natural language processing and machine learning were used in familiar products like Google Translate and other predictive text systems. These technologies analyzed data and made predictions based on patterns, but Gen AI introduces generative abilities that take the possibilities of AI a step further. 

Gen AI allows users to create new content—whether text, images, or even code—based on the data it’s trained on. This is a huge leap from traditional AI, which learns tasks by using results from large data sets (e.g. retrieving existing information). Users have gained familiarity with traditional AI systems over time and developed best practices, but the widespread introduction of Gen AI has sparked a new wave of both excitement and uncertainty as its capabilities push the boundaries of what AI can achieve.

Why lawyers should care about Gen AI

Gen AI is fundamentally changing the way that all knowledge work is done. Especially in the case of transactional law, which is inherently precedent-driven, there is a clear opportunity to leverage Gen AI technology to draft, review, and negotiate more efficiently. Law firm clients recognize this, and there is a growing expectation that firms utilize Gen AI to best serve their needs. 

Gen AI’s ability to generate entirely new text and ideas that mimic human creation allows lawyers to almost “converse” with their documents. This new functionality enables lawyers to engage with language-based information in completely new ways to optimize workflows and reallocate time for high-value tasks. 

While these new capabilities open doors to massive potential, without proper training, Gen AI tools create the risk of introducing errors. Comprehensive training to ensure that lawyers and firms understand the capabilities and shortcomings of Gen AI is more important now than ever before.

Real-World Examples of Gen AI

To be most effective in a transactional context, Gen AI should be tailored to specific sets of legal documents, like SEC filings, firm templates and precedent, or client-specific language. Lawyers and firms should understand how Gen AI can best be applied to optimize everyday workflows:

Expediting Summarization

We can see the value of Gen AI in the example of document or clause summarization. Often clauses or sections appear in large blocks of text filled with legalese. Having Gen AI summarize dense information in plain language can help lawyers quickly extract insights and move to the next stage of review. 

Enabling Document Q&A

Another example of the capabilities of Gen AI as applied to legal is document Q&A. Clients often ask attorneys for clarification on sticking points in an ongoing negotiation or in previous agreements. The ability to ask questions on a document or set of documents enables lawyers to be more responsive and informed in communications with peers and clients. 

Issue Identification

Gen AI's ability to quickly identify issues with a contract is another area of value. Gen AI can surface potential issues in a contract allowing lawyers to reallocate their time to analysis of the severity of the issues as they pertain to client goals. With less time spent on issue identification, lawyers can adjust focus to modifying contract elements to best protect clients and shift risk to the counterparty. 

Ensuring Ethical Application of Gen AI: Education is Key  

While Gen AI can achieve around 95% accuracy, it still requires careful oversight, especially concerning data privacy and confidentiality. Gen AI's responses are limited to the base of knowledge on which it was trained. This means that models applied to public data are susceptible to commonly held errors as well as information gaps. 

As Gen AI becomes more widespread, and products like ChatGPT and Claude are only a few clicks away, it’s essential that firms employ a systematic approach to educating lawyers on the proper and ethical use of Gen AI in their work. Inputting protected information on models applied to public data equates to publicly sharing information, constituting an ethical violation. Lawyers need to understand which environments are safe to share protected information and how to best access the benefits of Gen AI without violating client confidentiality. 

Oversight is another key lever to pull in the ethical use of Gen AI tools. AI-generated outputs should never be accepted without human oversight. Firms need to develop standards and processes to ensure the consistency of oversight, critically reviewing each result produced by Gen AI tools before sharing with clients. A systematic approach to oversight ensures the accuracy and maintenance of professional standards as Gen AI tools become more ingrained into everyday legal workflows.

Selecting a Data-Safe Gen AI Provider

When considering Gen AI tools for adoption, firms need to critically evaluate whether vendor technology complies with data privacy standards or creates the risk of sharing confidential information. 

The testing of and thorough evaluation of multiple solutions, along with a strong data governance policy, is critical for data safety. Undergoing a pilot period and getting hands-on time with a product also allows legal teams to better understand how Gen AI can impact their daily work and it’s ability to safeguard sensitive information.

At DraftWise, data privacy is a top priority considered in every step of our platform build. We do not access the data inputs that our clients provide, ensuring full confidentiality. 

Upskilling your team

The introduction of any new work process creates friction. To ease the adoption process, firms must employ a strong change management strategy to introduce and educate their lawyers on Gen AI.

Developing new habits and integrating technology into daily tasks requires time and effort on behalf of firms as well as individual users. The only way that lawyers can become comfortable with Gen AI is through repeated use and experimentation.

Innovation teams should encourage lawyers to experiment with Gen AI in a non-business setting to expand their understanding of how to interact with and extract value from Gen AI, as well as become familiar with its potential shortcomings. The initial uses of a tool are often the biggest hurdle to overcome - but with practice, lawyers can use Gen AI technology to help with tasks like contract revisions, creating issue lists, and handling open-ended questions. 

In recent years, it’s become common for firms to train summer associates in AI proficiency and prompt engineering – helping new lawyers understand from the onset of their work when and how to effectively use Gen AI tools. The understanding of ‘how can Gen AI help my work’ and ‘what risks using Gen AI can introduce into my work’ are fundamental to the successful integration of Gen AI tools into legal work at scale. 

Learn more about Mastering Gen AI

As Gen AI tools improve, understanding their capabilities is critical—generative outputs are reaching new levels of accuracy and consistency as tools become better at interpreting prompts. Now is the time to learn how to use Gen AI technology. 

Connect with the DraftWise team to learn more about the steps that firms and lawyers can take to leverage the value of Gen AI.

‍