Podcast: “Guardians of Code and Conscience: Exploring Legal and Ethical Frontiers of Generative AI”

If you missed the 29th Annual National Security Law Conference sponsored by Duke’s Center on Law, Ethics and National Security I have some good news for you as today we are continuing our posting of videos from the conference.

“Guardians of Code and Conscience: Exploring Legal and Ethical Frontiers of Generative AI,”  by my friend Brig. Gen. Linell Letendre not only hit it out of the ballpark, it took it to another universe. 

To categorize her presentation as spectacular doesn’t even seem to be a worthy description.  Countless attendees – including those with no legal (or science) studies praised her presentation.  In fact, I very much doubt you can find a better one-hour presentation on the legal and ethical issues of generative AI anywhere. It is the cutting edge of cutting-edge. Rare is a speech that so brilliantly weaves together scenarios that stir people to think and consider consequences, identify ethical issues, and provide historical and current context on a complex subject – all in an hour! This can elevate your understanding of AI and its use in the legal field.

General Letendre, who is now serving as the Dean of Faculty at the U.S. Air Force Academy, expertly explains the key aspects of generative AI in a way that is accessible to everyone.  She then weaves into the discussion the relevant provisions of the ABA Model Rule of Professional Conduct, along with examples and scenarios.

Seriously, this is amazing not just because of the superb substance (and what topic could be more timely?), but also because of the way she presents it.  Professors everywhere ought to take notes on this presentation as it is the gold standard as to how to teach complex topics. The video can be found here.  

My super talented research assistant, 3L Madison Cash (see here), wrote a reflection paper that outlines some of what you’ll hear and why this presentation is not-to-be missed.

As Brigadier General Letendre pulled up her Powerpoint and I read the words “Generative AI,” I prepared to be completely lost. I had recently read P.W. Singer’s “Burn-In: A Novel of the Real Robotic Revolution” and Andrew Krepinevich’s “Origins of Victory”, both of which discussed AI and new technological developments in warfare at a high level.

Although I had understood the basic implications and solemn warnings that often surround the use of AI, I still had a fuzzy understanding of how it actually functions and processes data.

Brig. Gen. Letendre

 In her talk, General Letendre made AI accessible to lawyers, which I think is vital to the legal profession moving forward. Without understanding the technology, we cannot hope to integrate it into our legal responsibilities and practices. Crucial to this accessibility was General Letendre’s ability to trace the history of the development of technology.

Too often, AI primers begin by discussing the most current, developed technology, without first explaining how those models developed. Because readers don’t understand the background technology, the functionality, development, and future potential of the new technology is more difficult to grasp.

First, General Letendre discussed classical AI, a model-based ruleset that operates similar to a chess game. She emphasized that the data was secondary to the ruleset in traditional AI, which explains why it often spits out overbroad results.

Next, General Letendre explained how AI has developed towards a machine learning model, which focuses on the type of data provided in order to drive its prediction.

She clearly laid out how important it is to structure a clearly delineated task to best guide the AI system and provide it with accurate and diverse training data—for the first time, because I understood the way AI developed, I understood how important it is to lay out the appropriate parameters of a task to get the best results.

Next, General Letendre moved on to generative AI, the current technological innovation taking the world by storm. Using the same foundational principles discussed at length involved in prior AI models, she demonstrated how generative AI uses vast amounts of data to predict the requested result; she also noted how neural layers can complicate and narrow the applicable parameters.

However, she was careful to note that the technology is not perfect and briefly discussed the reliance risks associated with generative AI that at times, hallucinates.

Hand in hand with her discussion of the risks was her legal analysis of the ethical obligations of an attorney to engage in the use of this technology. I thought it was so helpful that she applied the professional rules of ethics to the use of AI by presenting several variations of a case-study, teasing out the ways in which attorneys should be competent in using or directing the use of AI.

In the same way that an attorney must be competent in using a technology, they also have a supervisory role over junior legal personnel, and even the technology itself, to best accommodate the client’s interests.

Use of the technology thus requires a basic understanding of the technology, how to maximize its use, and how to double and triple-check its results. However, this extra work to both understand and verify the technology is worth it–as General Letendre concluded, “AI will replace lawyers unless lawyers begin to use AI.”

To hear all of General Letendre’s fabulous presentation from the LENS conference, be sure to check out the video found here. Believe me, it’s a mistake if you don’t watch this one!.

My fireside chat with CIA General Counsel Kate Heinzelman from the Center on Law, Ethics and National Security’s 29th conference is also now online and can found here.

Unless otherwise indicated, Conference speakers are expressing their personal opinions, and not necessarily those of their employer (to include the U.S. government), the Center of Law, Ethics and National Security, Duke University, or any other person or entity (see also here)

Remember what we like to say on Lawfire^®: gather the facts, examine the law, evaluate the arguments – and then decide for yourself!