Given the rapid expansion and seemingly endless uses of AI-generated writing tools, Crafty Counsel has been in touch with in-house lawyers in our community to ask them how they’re going about using programmes like ChatGPT and what advice they can give to those who are just dipping in the AI pool. 

Written by Emma Haywood, Director and Principal Consultant at Bloomworks Legal

How have you been using GPT-4?

Like most, I’ve started with ChatGPT as a gateway to understanding the scope of generative AI and potential use cases for lawyers and beyond. I’ve focused on building the right skills to get the most out of the technology, including diving into the fascinating world of “prompt engineering”. Most people’s first instinct is to use ChatGPT and other conversational large language model tools as a search engine, providing brief keyword-based prompts with little contextual detail. But that approach doesn’t really take full advantage of what these models can do. 

I’ve been experimenting with layered or phased prompts, because I’ve found that they can generate much more targeted and useful outputs. For example, if I need to refine my strategic thinking on an issue affecting the business I’m supporting, I might begin by asking ChatGPT to provide an outline of a helpful general concept or framework: “Describe the main principles of change management theory”.  I’ll then refine that general outline: “Re-write the response above to focus on the aspects that would be of most interest to a senior lawyer advising c-suite executives within a growing tech company”

Once I’m happy with the outline, I’ll prompt ChatGPT to apply that outline to a relevant scenario: “How could the above principles be used when introducing a new process for evaluating and approving new commercial partnerships within a business? Provide practical tips as to how to engage with stakeholders and successfully embed the new process”. This is a useful method for super-charging my strategic thinking when solving legal and business issues. It’s a bit like a free MBA for GCs.

Beyond that, I’ve signed up for demos and trials of a few GPT4-powered legal tech tools, such as Co-Counsel (which is currently US-only), JosefQ and the AI assistants offered by existing contract management platforms. While it’s important to have a birds-eye view of the more general tools that might be of interest to your colleagues across the business, it’s hugely beneficial for in-house lawyers to engage with legal-specific platforms as part of the learning process. Typically the legal tech providers have a deeper understanding of the concerns we have as lawyers (whether that’s confidentiality, IP, legal privilege or simply a higher-than-average level of interest in the user terms and conditions), and it’s helpful to get a practical insight into lawyer-friendly use cases without any long-term commitment.

What would be your advice for in-housers who want to start using GPT-4 but don’t know where to start?

As with any legal tech or legal ops project, it’s best to start with the problem you want to solve, rather than trying to find a way to shoehorn a new tool into your tech stack. 

First, identify a few low-risk, low-value tasks that you’d like to streamline, eliminate or knock off your to-do list. These might be refining the questions on your team’s legal intake form, preparing guidance materials and FAQs for business colleagues, summarising the key points of a new regulatory update for a board presentation, or testing your thinking when updating provisions in your standard contracts. I like to think of generative AI tools (whether ChatGPT or legal-specific platforms) as a helpful but relatively inexperienced assistant that frees up your time to focus on higher-value, strategic work.

You’ll need to exercise a degree of caution when using generative AI, but that shouldn’t stop you from experimenting in an informed way. After all, as lawyers, one of our superpowers is balancing risk against business objectives. There’s also an inherent risk in doing nothing, sticking your head in the sand or telling yourself you’re too busy. If you’re using publicly available tools or beta versions of enterprise tools, check the terms carefully, especially when it comes to confidential information, IP and personal data. All output should, of course, be checked and overlaid with your own judgement and expertise. 

What are you planning to do next?

I’m conscious that the generative AI landscape is constantly changing, with new tools, regulatory developments and court cases emerging all the time. Keeping up to date is almost a full-time job, so to avoid the overwhelm I’ve curated a small collection of valuable resources that I continue to delve into whenever I can. My top recommendation is The BrainyActs, a daily email newsletter written by Josh Kubicki, a US-based legal innovation professor. It’s practical and interesting for in-house lawyers, but also brings in ideas from outside the legal sphere.

Finally, I plan to explore how generative AI could improve working life for lawyers, beyond the obvious productivity boost. One of my passion topics is the 4 day working week. As well as writing about the interaction between generative AI and alternative working patterns via my newsletter, The 4 Day Lawyer,

I’ll be on stage in the Marquee at Crafty Fest on 14 June talking about how we can harness the power of innovation (including generative AI) to support work/life balance for in-house legal professionals. See you there!