At MIT, Martha Minow speaks on sustaining democracy, both by using and overcoming AI

Harvard legal scholar Martha Minow used MIT's fall Compton Lecture to argue that artificial intelligence is eroding American democracy but could also help shore it up, MIT News reports. Speaking in Huntington Hall after an introduction by MIT President Sally Kornbluth, Minow, a constitutional law and human rights expert and former Harvard Law dean, framed her talk, "Constituting Democracy in the Era of AI," around four requirements for self-government: education, reliable information, security from violence, and fairness. She cited Gallup polling in which 77 percent of respondents said the nation's founders would be disappointed in the country today.

Much of her talk addressed schools, where she said funding gaps and growing economic and ethnic segregation are compounded by AI tools that encourage passivity. She warned that sycophantic models threaten young people's capacity for memory and critical thinking, and urged her audience to build tutor-like systems that prompt students with hints and questions instead of doing their work, along with assessments that are harder to cheat on.

Minow also said modern media technology has contributed to a 75 percent drop in journalists in recent decades, with corruption rising where local coverage has vanished, and proposed that technologists create shared "information commons." On security, she said every judge she knows has faced a credible threat of violence.

Despite her fear of losing Enlightenment commitments to reason, evidence and tolerance, Minow maintained that a turnaround is possible if researchers devote their talents to it. The talk matters as a call for technologists themselves to take on AI's civic costs.