AI, Ethics, and Public Service: Bridging the Gap through Stanford HAI’s Fellowship Program

As artificial intelligence becomes ubiquitous, impacting the lives of people with and without their knowledge, demand is growing for federal policies to promote innovation and oversee its use. Yet those most capable of understanding the complex challenges surrounding emerging technology are often missing from critical high-level discussions.

“The key necessity in developing AI policy, especially for government policymakers, is being able to access technical talent,” says Daniel Zhang, senior manager for policy initiatives at Stanford’s Institute for Human-Centered Artificial Intelligence (HAI). “You simply can’t invest in or regulate AI without an expert in AI. But the data from our AI Index shows that less than 1% of all new AI PhDs graduating in North America go into government work.”

Learn more about the Tech Ethics & Policy program and apply here. 

 

In an attempt to reverse that trend and boost the number of AI experts entering public service, Stanford HAI and the Ethics, Society, and Technology (EST) initiatives at Stanford’s McCoy Family Center for Ethics in Society joined forces in 2022 to launch the inaugural Tech Ethics & Policy summer fellowships, chaired by HAI Senior Fellow and professor of political science Rob Reich and HAI International Policy Fellow Marietje Schaake. Graduate fellows are required to have taken at least one course in computer science or symbolic systems or multiple courses in the School of Engineering and enroll in the spring fellowship course “Introduction to Tech Ethics and Policy Career Pathways,” taught by Reich and Schaake, before being matched for summer jobs at executive branch offices, congressional offices, and think tanks in Washington, D.C.

“Our goal is two-fold,” Zhang says. “To supply the government with AI technical talent and to allow students to experience what a public service career could look like in the hope more will opt for careers in government. Governments don’t do much recruiting on the ground here in Silicon Valley, and the temptation to go into private industry is strong, but we know there are a lot of Stanford students interested in public service. This is a bridge to another possible path.”

The fellowship program includes an undergraduate track (run by the McCoy Family Center). Applications are now being taken for both options for the 2023-24 session. The fellowship program, Zhang says, will hopefully contribute to the call for an “AI Talent Surge” to recruit more technical talent into the federal government, which is included in the October 30 White House Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.  

Avi Gupta (MS, computer science), Liana Keesing (BS/MS, electrical engineering), and Regina Ta (BS, symbolic systems and comparative literature; MS, computer science) were among the program’s inaugural nine fellows. Here, they describe their summer experiences and what future fellows should expect from the program.

Avi Gupta: Drafting a White House Executive Order on AI

Gupta has long been interested in the intersection of technology, policy, politics, and law. He spent his fellowship at the White House Office of Management and Budget (OMB), where he was tasked to work on developing a draft policy for how federal agencies manage the risks associated with AI. In addition to responding to congressional inquiries, Gupta was able to observe firsthand the complex interaction of White House offices, internal stakeholders, and government agencies.

“A big part of what I was doing was helping write guidance to assist federal agencies in implementing AI safely and responsibly. It gave me the opportunity to talk with pretty much every major agency, which was really amazing,” Gupta says.

During this time, however, his original job description unexpectedly changed.

“The impacts of generative AI made this work an administration priority and resulted in an executive order that came out on October 30,” he says. “Because the small team I was on had expertise in AI, I ended up playing a role in drafting that executive order. It was amazing. I was in the right place at the right time.”

The fellowship program provided opportunities Gupta never anticipated.

“Getting to have my work reflected in policy that’s going to shape the federal government’s approach to AI was an incredible experience and a tribute to the uniqueness of this opportunity and how well my work at Stanford prepared me to contribute to that conversation,” he says. “I’ve long been interested in public service, and for this fellowship I wanted to learn how executive branch policymaking worked at an agency or in the administration. I never imagined I would get to do this.”

Liana Keesing: Rethinking Federal Procurement of AI

Keesing grew up near Washington, D.C., in a family of technologists, her interest in engineering rivaled only by a deep fascination with the democratic process.

“I grew up knowing that government is not a monolith and that small decisions made every day have a tremendous impact on the country,” she says. “I also think that the most interesting thing about an engineering education — something I wish we valued more in policymaking and government — is that it’s excellent training in how to solve all kinds of problems.”

Keesing spent her summer fellowship working at the Senate committee on Homeland Security and Governmental Affairs, chaired by Michigan Sen. Gary Peters. On her first day of work, knowing the committee regularly scheduled AI-related hearings, she made a suggestion.

“I’m particularly interested in federal procurement policies for AI, so I made a case for a hearing on it, and they agreed,” she says. “The government is the largest buyer in the world, but AI is so new that there really aren’t yet any industry standards. The decisions the government makes have the potential both to become industry standard and to create huge ripple effects, as Microsoft and other companies adjust their focus to provide those things.”

Keesing and her team worked with government representatives, trade groups, academics, business leaders, and ethicists to prepare for the hearing. Today, she continues to work for the committee as they use the information gleaned from that hearing to craft the framework for a new bill they hope will be introduced into the Senate dealing with the federal procurement of AI.

“We are really in the thick of these discussions, and I hope in the future there’s an increasing push to have more people here on the ground who have engineering as the primary thrust of their career,” she says. “I always knew I wanted to work in tech policy in D.C., but I didn’t know I’d be part of this program with other fellows and a network of connections, building a foundation for the future. I think there’s an appetite for this on campus, and it’s a model that means students end up working in really exciting places.”

Regina Ta: Building the Conversation Around Generative AI

Ta, interested in entering policy from a research perspective, joined the Brookings Institution think tank for her summer fellowship.

“I was placed at its Center for Technology Innovation, where my assignment was mainly as a tech ethics and policy fellow,” she says. “Right off the bat, I felt very integrated on the team, because there was a lot of room for me to take initiative on projects or ideas that piqued my curiosity.”

Among the work Ta undertook at Brookings was a paper she pitched and developed on how schools were approaching generative AI governance in classrooms. Another dealt with generative AI, linguistics, and marginalized speakers. Both articles were published on the Brookings website.

“When I had an idea, I was able to take ownership of it, pitch it, and go out and seek co-authors among the senior fellows on my team,” Ta says. “My co-authors continued to collaborate with me, even after I left D.C. and the fellowship, to get them published, which was amazing.”

Ta, who enjoys producing podcasts, was also able to work on the Brookings “Tech Tank” podcast throughout the summer, eventually initiating and helping produce an episode on the implications of new moon landings for the future of space exploration.

Her summer in Washington, Ta says, taught her the role think tanks play in promoting conversations and interacting with federal agencies, various organizations, and industry.

“Across all this collaboration, you see where people’s incentives start to align and where there’s room to drive policy,” she says. “Coming to D.C. imbued me with a sense of possibility that there’s an intersection to be found here and there are collaborations to be made. So, did it make me want to work in and with government public policy and adjacent areas for the public and civic interest? Definitely yes.”

Applications for the next Tech Ethics & Policy fellowship are due Dec. 1, 2023. Learn more and apply.

Stanford HAI’s mission is to advance AI research, education, policy and practice to improve the human condition. Learn more.