About That FIRE Study

The reporting on the new Foundation for Individual Rights and Expression College Free Speech Rankings focuses on how things haven’t changed. The headline of Johanna Alonso’s excellent piece is “Students Report Less Tolerance for Controversial Speakers.”

To be clear, the issue of tolerance for campus speakers—and the physical safety of speakers and attendees—remains paramount, as last week’s violence made clear. But for me, FIRE’s study misses the single most profound change on college campuses: AI, and the reality that students are increasingly doing their intellectual exploration privately, not publicly. FIRE’s survey doesn’t ask what questions students are asking their AI models in the privacy of their dorm room or quietly on their laptops during lectures. Yet Inside Higher Ed’s recent survey makes this clear: “Most students are using generative AI for coursework, but many are doing so in ways that can support, not outsource, their learning.” As one student put it, faculty members and leaders “need to understand how accessible and potent it is.”

FIRE’s approach conflating intellectual freedom (the right to speak) with intellectual obligation (the expectation that one should speak) belongs to an older era. The study’s methodology treats public expression of controversial political views as an expected norm rather than a choice complicated by AI and social media. Students who report discomfort expressing their views in classroom discussions, campus common areas or on social media are counted as evidence of a “chilling climate,” regardless of whether these students have any desire to engage in public discourse.

In the AI era, public self-censorship might be better seen as “I’m having these conversations with my chat bot, thank you,” rather than problematic restraint in sharing controversial views. Even before AI, I was concerned that measures of the freedom to speak assume that students ought to speak. If the ubiquity of AI in universities should change one thing, it should be how we approach intellectual liberty.

The evidence is everywhere that students prefer to explore viewpoint diversity from their AI models, rather than from “speakers.” Scott Latham, predicting a very different (but very likely) future of “the AI university” in The Chronicle of Higher Education, foresees prospective students specifically inquiring about AI-taught classes and parents asking, “Does your institution have AI advisers and tutors to help my child?” Students are already expecting AI-mediated learning rather than traditional classroom discussions. The IHE survey noted that “Generative AI is influencing students’ learning and critical thinking abilities.” Yes.

The FIRE study operates as if we still live in an era of scarcity of views, where controversial voices are rare and need protection or positive platforming. The FIRE study does not mention AI and so cannot conceive that students are reading about every single controversial subject possible—gender transitions, income redistribution, progressive taxation, cryptocurrencies, euthanasia, eugenics, you name it. Students can summon any perspective on demand without the social costs that the survey obsessively measures. While professors worry about chilling effects in classroom discussions, students may already be engaging in far richer intellectual exploration through private conversations with AI systems. The IHE study reported that “performance pressures … are driving cheating” and that students “reject policing.” I see ongoing confusion about a definition of cheating that could simply mean exploring ideas unauthorized by a faculty member, without surveillance.

In short, the FIRE study’s emphasis on public expression ignores this new reality where new forms of intellectual exploration happen in ways the survey does not even recognize. Students’ reluctance to perform controversial viewpoints in public settings may reflect a paradigm shift that meaningful engagement with controversial ideas requires privacy and reflection. AI may deliver all this far more fruitfully than the performative classroom discussions that FIRE champions.

Consider all the ways the study’s methodology is outdated in the AI era. FIRE measures “comfort expressing ideas” by asking students about “expressing your views on a controversial political topic during an in-class discussion” and “expressing your views on a controversial political topic to other students during a discussion in a common campus space such as a quad, dining hall, or lounge.” These questions assume that intellectual engagement requires public performance. The survey treats discomfort with these activities as problematic, asking, “How often do you self-censor during conversations with other students on campus?” and defining self-censorship as “refraining from sharing certain views because you fear social … professional … legal … or violent consequences.” I see no recognition of how students are using AI. The performative nature of campus discourse that FIRE measures ignores private spaces.

The rankings methodology compounds this flawed thinking by penalizing institutions where students report discomfort. Colleges receive failing grades when students avoid public political discussions, as if intellectual health could be measured by willingness to perform controversial opinions in social settings. The study notes that “166 of the 257 schools evaluated received an overall score below 60—earning a failing grade for their speech climate,” treating what we might call discretion (or disinterest) as institutional failure.

Yet, as Latham predicts, by 2030, “incoming students will also be given a personalized AI agent” that can “construct a course that transcends the classroom, campus, and time.” Students will work with AI agents to “find expert scholars across the globe, line up real-time or recorded video lectures, and simultaneously incorporate material from YouTube, Google, and university libraries.” If students are already using AI to write their papers and do their homework, they are also engaging with controversial ideas, testing arguments, considering counterpoints and along the way developing their thinking without professional consequences or social judgment.

In this context, the study’s focus on “tolerance for controversial speakers” represents a fading ideal. Students who can access any perspective instantly through AI have far less need for campus speaker events. When FIRE measures whether students think their colleges should “allow or not allow a speaker on campus” who holds various controversial positions, it assumes that human speakers remain the primary vehicle for intellectual diversity. This assumption belongs to an earlier era. I confess I’m a little sad to see it go, but there it is.

Even forward-thinking educators acknowledge this shift. Latham notes that when one colleague demonstrated using AI for course design, another “leaned over and asked, ‘Then what’s our job?’” While faculty grapple with their changing role, FIRE continues measuring an antiquated model of intellectual engagement.

For today’s students, a higher percentage of learning may be happening privately, in spaces protected from social surveillance. Their apparent reluctance to perform publicly may signal intellectual maturity rather than institutional failure. FIRE’s rankings system, built for a world before AI, does not recognize this transformation. They’re still measuring all the wrong things in the wrong places, missing entirely how intellectual diversity circulates in an age of synthetic voices and private digital exploration.

Finally, last week’s assassination of a speaker on a university campus makes the name “FIRE”’s direct association with the act of shooting a gun impossible to ignore. As some faculty and staff are fired for their speech following the shooting, the name simultaneously represents the very consequence the organization fights against. The acronym now stands as a constant, jarring reminder of violence, professional ruin and the most famous legal exception—what not to cry in a crowded theater—to the rights FIRE purports to champion. It is worth thinking about.