DHS’s New ‘AI Playbook’ Ushers in ‘Tomorrow’s Lawyers’
And some thoughts for legal education today
This week, DHS released its Playbook for Public Sector Generative Artificial Intelligence Deployment. In this “AI Playbook,” DHS outlines (mostly for other federal agencies) the best practices it has gleaned from three pilot projects to aggressively develop its AI capabilities.
In those three pilot projects, announced in March 2024,
· Homeland Security Investigations (HSI) used a large language model (LLM)-based system to more quickly and accurately search public documents for things like fentanyl networks and child exploitation perpetrators;
· The Federal Emergency Management Agency (FEMA) used an LLM-based system to assist communities in creating hazard mitigation plans;
· United States Citizenship and Immigration Service (USCIS) used generative AI to train refugee and asylum officers on the law according to the officers’ specific needs and allow them to practice interviewing skills on their own, as often as needed.
The “Ideal” Future Agency Staffer
These pilot projects are interesting but, like most of AI, still infant. And you probably don’t run a federal agency, so why would you care?
I don’t run a federal agency either, but I do train law students (and I know some of you reading this do too). What jumped out at me was this statement, “play” number six on page 14, titled “Training and Talent Acquisition”:
Employees who use any GenAI application should understand its capabilities, limitations, risks, and opportunities. Managers who may oversee GenAI use cases should have a baseline familiarity with the technology to ensure they take an informed approach to their respective roles and responsibilities. Organizations should, therefore, generally train their staff on the use of GenAI tools and help them develop fundamental technical skills. (Emphasis in original.)
Step one, the Playbook says, is for organizations to identify necessary technical skills and see if anyone on their team already has them. If not, they should, “where possible,” create training that allows “current employees targeted opportunities for up-skillinng or cross-training.” But they should also “consider how to hire skilled employees.” (Emphasis, again, in original.)
Tomorrow’s Law, and ‘Tomorrow’s Lawyer’
This may not seem like a big deal. Sure, train your employees and hire some technical support staff. That’s pretty much the same old playbook that every law firm and legal organization has followed since we started using personal computers, right?
Maybe not. In his visionary book, Tomorrow’s Lawyers: An Introduction to Your Future, Richard Susskind describes the transformation of the legal system, and by extension the work of lawyering, that the coming technological revolution will instigate. Technology may free the average person or business from needing one-to-one consultation with a lawyer to comply with regulations or resolve disputes, Susskind writes, because technology will increasingly embed legal standards into our everyday environments. He gives the example of Smart Buildings that will be designed to meet environmental standards automatically. (In the immigration context, might we someday see Smart Borders that automatically detect, screen, admit, or divert applicants to law enforcement personnel?)
According to Susskind, tomorrow’s laws will still require lawyers, but those lawyers will need different skills than case analysis and oratorical flair. Rather than providing one-to-one advice, tomorrow’s lawyers will be “legal knowledge engineers,” “legal technologists,” “legal operations experts,” “legal process analysts,” etc. In other words, legal knowledge will still be needed, but the role of the legal expert will be to engineer and analyze how legal standards are embedded, operated, managed, and updated within our technologically built environment.
DHS’s AI Corps and the Advent of ‘Tomorrow’s Lawyers’
Reading DHS’s AI Playbook in the light of Susskind’s predictions, the report’s sterile agency-speak struck me as code for something much more radical: the initiation of tomorrow’s lawyer within federal agencies today.
Sure, agencies right now would rather train up their existing staff on nascent technology than hire thousands of new people with highly in-demand skills. (DHS did, though, initiate an “AI hiring sprint” in 2024 that led to the rapid hiring of thirty-one technology experts into what they dubbed their “AI Corps.”)
But Play #6 suggests that DHS already appreciates that those who employ AI technology in service of federal government operations – even line level managers – should have a baseline understanding of how that technology works. Anything less would leave the federal government unprepared to deploy its regulatory apparatus effectively or to respond to public challenges over accuracy, privacy, and civil liberties.
The days of the lawyer simply calling the IT department – a conversation from which both walk away frustrated because they don’t speak each other’s languages – will soon end. No longer a mere add-on to the knowledge work that lawyers have done for centuries, technology is rapidly merging with law, and will demand a workforce trained to understand and manage they hybrid.
Lessons for Legal Educators and Law Students
Legal education is unprepared for this challenge. The NextGen Bar Exam is only just now coming around, moving from testing applicants’ recall of thousands of legal doctrines they’ll probably never work with – a task that has been obsolete at least since the advent of legal search engines – to some form of legal skills testing. I fear that the preparation of ‘tomorrow’s lawyers’ may be inaccessible within the slow-moving institution of legal education.
The best bet, for my money, is on the courses in entrepreneurship for lawyers that a handful of law schools now offer (including a pilot last spring at WVU Law). By developing an entrepreneurial mindset, students now completing a traditional legal education can learn to approach the massive shifts in the law and the profession with the flexibility and resilience that every entrepreneur must cultivate.
Even better, young and aspiring legal professionals may want to pursue heretofore unconventional companion degrees in information technology, computer science, technology management, systems engineering, or business entrepreneurship.
Not technologically inclined? You soon won’t have any choice – train yourself or count on your employer to continue to expend the resources to train you.
Still not convinced? There’s always English literature. (Not that there’s anything wrong with that.)