Shaping the Future of AI

Artificial Intelligence (AI) seems to be developing at warp speed. And Sneha Revanur ’26—whom Politico calls “the Greta Thunberg of AI” and who was named the youngest member of the Time 100 Most Influential People in AI—is working to ensure that we all stay one step ahead.

In high school, Revanur founded Encode Justice with the aim of “fighting for human rights, accountability and democracy in the age of AI,” according to the organization’s website. Its membership has since grown to 900 high school and college students around the globe.

Revanur grew up in California but is becoming increasingly familiar with Washington, D.C. This past summer she had an internship there at the Center for AI and Digital Policy, a nonprofit research group, with the support of Williams’ Alumni Sponsored Internship Program. And in October, she was on hand when U.S. President Joe Biden signed an executive order regulating AI.

Back on campus, Revanur has been busy running her nonprofit and being a sophomore. “It’s a lot of juggling calls in Sawyer and working to get papers done,” she says, adding that she’s grateful for her professors’ and peers’ support.

“I really want to be involved in this conversation long-term,” she adds. “My highest-order desire is that I keep doing this work until I don’t need to do it anymore because it’s not necessary.”

Here’s where she’s focusing her efforts.

RESEARCH: One of Encode Justice’s first campaigns was “studying the impacts of a recidivism prediction tool that had been used by jurisdictions across the country to determine whether or not a defendant is likely to re-offend,” Revanur says. “We found staggering racial bias in that tool as well as other sorts of tools that are used in hiring and education and everyother zone in public life. It’s really important that we break free from this perception that algorithms are automatically objective and neutral and think more critically about where we are deploying them in society.”

AWARENESS: Revanur and her colleagues are offering workshops to students around the world on topics including the ethics of AI, the intersection of social justice and computing, and algorithmic bias. More than 15,000 students have participated to date. 

Encode Justice is also focused on the 2024 elections, raising awareness about the misuse of generative AI, which, by “learning” patterns from images, text and other media, can produce strikingly similar content. That content can be used to “pump out misinformation,” Revanur says, potentially influencing voters.

“Because there is a veneer of objectivity,” she says, news consumers have a hard time distinguishing fact from fiction. “I’m really concerned about whether we’re potentially heading into something that has been described as reality collapse—a point at which we don’t know what to believe and we don’t have a sense of reality anymore because AI has transformed that notion of reality to an unrecognizable degree.”

ADVOCACY: Revanur and Encode Justice helped draft the Blueprint for an AI Bill of Rights—five principles guiding the design, use and implementation of automated technology—released by the White House in 2022. Last May, the group sent an open letter to the White House Office of Science and Technology Policy seeking the inclusion of young people in congressional leaders’ AI oversight and advisory boards. Revanur says Encode Justice is planning to endorse whichever candidate in the 2024 presidential election has the strongest platform in terms of AI regulation.

During her internship, Revanur took part in a roundtable discussion about AI convened over the summer by Vice President Kamala Harris. She talked briefly with Lina Khan ’10, head of the Federal Trade Commission, and met with staffers for U.S. Rep. Don Beyer ’72, a member of the House Committee on Science, Space and Technology. Each is a “huge player” in AI, Revanur says, adding, “I’m so filled with pride every single time I see Williams alumni out there influencing the conversation.”

INTERCONNECTION: Revanur says that it’s possible to rein in AI, but “it’s going to require a lot of intergovernmental coordination, because AI development is becoming increasingly borderless.” Encode Justice’s international chapters are busy pressing political leadership in their countries to take action.

“We’re going to need buy-in from rivalrous states and people who might have competing interests,” she says. “Unless we have a harmonized global AI governance approach and some clearcut, standardized codes and regulations for AI development, it’s going to be difficult to enforce. As important as our work is here on the U.S. congressional front, it’s only going to make a difference if we’re tying that up with action on the international front to ensure intergovernmental coordination.”