A Sneak Peek at the AI & Chatbot Summit
August 25, 2023
No other technological advancement in 2023 has gripped the professional world with equal fear and fascination than AI and ChatGPT. On September 18, the D.C. Bar CLE Program will host the Artificial Intelligence & Chatbot Summit to examine how attorneys use AI and open-source tools in their day-to-day practice and the legal and ethical issues involved.
Faculty expert Nicole Black, senior director of subject matter expertise and external education at MyCase, will help lawyers navigate the pros and cons of using this technology with her presentation “AI & Chatbot: An Overview.” Here, Black offers a preview of what she will cover during her session.
Talk about your upcoming presentation. What will attendees learn from it?
I’ll be providing an initial overview of [what] artificial intelligence (AI) and generative AI are, why they’re important, and why lawyers need to understand them. I’ll explain how there’s an exponential rate of technological advancement happening right now and why technology competence really requires lawyers to jump on board and learn as much as they can.
Then I will do a little bit more of a deep dive into some of the issues lawyers need to think about when they use this type of technology — [implicit] bias, professionalism in terms of ethics, issues regarding [client privilege], confidentiality. And then I’ll provide at least 10 different ways lawyers can safely use ChatGPT and generative AI, given the state of the technology right now.
ChatGPT has pretty much dominated the news in terms of emerging technology. What other AI-related developments should lawyers have knowledge of?
Essentially, it’s all generative AI. That’s really the one that lawyers need to stay abreast of. I’ve been presenting and writing about artificial intelligence for at least five years now and tracking the more traditional artificial intelligence tools. I’ll cover those categories during the presentation.
But generative AI really is a game-changer. So, it’s not necessarily just ChatGPT but generative AI as a category. I would say the legal technology companies that are coming out with guardrailed generative AI tools within their systems are probably the best way for lawyers to adopt this tech.
Part of what you’re seeing is the broader categories of artificial intelligence that were around before generative AI started getting built into them. So, it’s really becoming a part of those more traditional categories, which include legal research, contract analytics, contract drafting, document drafting, and litigation analytics.
What safeguards would you recommend for any lawyer or law firm using a chatbot or ChatGPT to ensure they are getting the most up-to-date and accurate information?
Well, first we have to determine whether they’re using a consumer generative AI tool like ChatGPT or … [one] built into legal software. If they’re using ChatGPT, Google’s Bard, or Bing Chat, then they need to think about confidentiality and avoid inputting any confidential information, making sure that their settings are, if possible, set so that their input isn’t going be used for training purposes. But then, you still run the risk of confidentiality issues with those more general consumer-facing tools.
So, you want to avoid inputting confidential information, and you need to have a basic knowledge of the topics you’re asking about so that you can vet the answers and determine whether they’re accurate. I would suggest avoiding using the general [ChatGPT platforms] for legal research at this point. Use them more for summarizing documents that you provide to them that have been edited or redacted to [remove] confidential information. Have them bullet-point documents, draft templates that you can use — letters, emails, or other correspondence — then make them more specific to the case. Those are the ways I suggest using the more general consumer-facing products.
But, if you’re talking about generative AI tools that are based on GPT power tech and are built into legal software, then you can — based upon the assurance that the company gives you about how they’re handling your firm’s data if you’re not comfortable inputting confidential information — proceed, while understanding how they’re guardrailing it in terms of inaccurate results. Are they telling you there may sometimes be some? So, you still need to review the work carefully. And regardless of what tool you use, [for] anything that you submit to the court or that you’re relying on for client representation and advice, you always have the obligation to review it for accuracy before submitting it.
You mentioned that generative AI technology has been a game-changer. It has also invoked a lot of fear across multiple professions that this technology will eliminate a lot of knowledge-based jobs. Others, however, argue that new careers may emerge. What is your take on that regarding the legal profession?
I think both of those assessments are correct. Just like with other technology, jobs are going to be replaced, or parts of jobs are going to be replaced, depending on the functionality you’re talking about. But that’s already happened with other types of software, too. I think it’s going to happen more quickly and on a slightly broader scale with generative AI software.
But I also think that it’s going to create more jobs in terms of harnessing the technology, understanding it, and educating lawyers about it. It’s going to change what the base level of work the people do in law firms looks like.
And, ultimately, it’s going to affect the billable hours, which everyone’s always predicted. It’s going to encourage lawyers to approach pricing [differently by] looking at how much time it takes to create documents, how much time a lawyer needs to interact with their support staff, and how long it actually [takes] to get the work done.
It’ll be interesting to see, and it’s really difficult to predict. But, at the end of the day, we’ll absolutely get rid of a lot of the mundane work by streamlining [many] tasks. It’s going to change a lot of what the legal field looks like, but not necessarily for the worse. We’ll see.