AivexaNewsSearch
AI news for builders and product teamsChecked every hour

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

Collected Oct 23, 2026

Martha Minow, the 300th Anniversary University Professor at Harvard University, delivered MIT's Karl Taylor Compton Lecture on a Thursday this fall, speaking on "Constituting Democracy in the Era of AI" in Huntington Hall, Room 10-250. MIT President Sally A. Kornbluth introduced her, calling Minow a legal scholar of extraordinary intellectual scope and a dedicated educator. Minow, a faculty member at Harvard Law School since 1981 and its dean from 2009 through 2017, opened with a bleak assessment of the present: it is not a good time for democracy or constitutionalism.

She listed the strains she sees on U.S. governance, among them violations of laws and norms, stifled free speech, diminished congressional authority, disregarded legal rulings and the routine vilification of others. She cited recent Gallup polling in which 77 percent of respondents said the country's founders would be disappointed in its current condition, and said many people genuinely wonder whether the system of government and rule of law, ignited 250 years ago, will survive.

Minow organized the talk around four ingredients of democracy: education, reliable and accessible information, security from violence, and fairness. Her method was to take each in turn and ask how AI bears on it, then to propose ways the technology could be turned toward repair. Her stance was not resignation; she argued a turnaround is genuinely possible if people at MIT and elsewhere spend time and talent on solutions.

Education received the most attention. Minow noted that U.S. school funding already varies sharply from place to place and that schools have grown significantly more segregated by economics and ethnicity over the past three decades. Against that backdrop she argued AI risks further weakening independent thinking at exactly the moment civic understanding matters most. She described dominant AI platforms as intensifying the techniques social media companies already use to induce passivity, and singled out sycophantic model behavior as a danger to young people's capacity to think, remember and reason critically.

Her proposed remedy is design work: AI that behaves more like a tutor, offering nudges, questions and hints instead of producing finished work. She said existing large language models, training methods and interface design could already support this, but that much more could be built to be trustworthy and to deepen curiosity rather than substitute for people. Addressing the MIT audience directly, she said those present could build better tools than what exists and devise assessments that are more meaningful and resistant to cheating.

On information, Minow pointed to the civic cost of misinformation and of the shrinking press. She observed that modern media technology has coincided with a 75 percent reduction in journalists in recent decades, and that corruption rises where local media disappears. Her suggestion for civic-minded technologists was to build a widely used information commons, or several of them, that helps people reach more reliable material. In practice, a shared information commons is a hard engineering and governance problem: someone must decide what counts as reliable, keep the corpus current, and prevent the same distribution channels that spread falsehoods from degrading the new resource. Minow's framing treats it as a design and stewardship task rather than a purely technical one.

On security, she was blunt. She said the country is on the brink, that she knows many judges and every one of them faces a credible threat of violence, and that this is not acceptable. On fairness she examined several dimensions of unequal treatment in American life, treating equity as a load-bearing element of democratic legitimacy rather than a separate concern.

In a Q&A after the remarks, with Kornbluth reading audience questions aloud, Minow named her deepest worry: the loss of the Enlightenment. She described it as the belief that reason, hypothesis testing, the rejection of myth in the face of evidence, toleration, rights and wider participation can produce a better life, and she connected that concern directly to disinformation and to the erosion of critical thinking in schools. She noted that much of her career has dealt with mass violence and what happens when people misunderstand and fear one another.

The lecture series itself dates to 1957 and is MIT's highest-profile speaking event. Kornbluth recalled Karl Taylor Compton, MIT's ninth president from 1930 to 1948 and later chair of the MIT Corporation from 1948 to 1954, as a leader who carried the Institute through the Great Depression and World War II and turned it from a training school for job-ready engineers into a global research university, and as a committed patriot who helped mobilize university research for the national effort.

The talk offers no product announcements or benchmarks, so its practical value for builders lies in the design agenda it sets out. Minow's argument implies that tutoring interfaces, refusal and uncertainty behavior, and assessment tools are civic infrastructure, not just consumer features. A likely trade-off is that a genuinely tutor-like model, one that withholds answers and asks guiding questions, will feel less immediately satisfying than a compliant assistant, and that usefulness of this kind is harder to measure than engagement. Her call for cheating-resistant assessment also points at a gap product teams rarely address: evaluations that test reasoning rather than recall.

Why it matters: Teams building education, media and civic tools are being asked, from a prominent legal and human-rights voice, to treat sycophancy and answer-giving as harms rather than conveniences. For developers of models and interfaces, that suggests measuring whether users think better, not just whether they get answers faster, though whether such metrics can be standardized remains an open question.

Read at MIT News · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

In Compton Lecture, the distinguished legal scholar warns of potential “loss of the Enlightenment” while suggesting ways innovations can meet our civic and cultural needs.