Student Surveillance Software: Speech First v. Virginia Tech
Suppose a university has a website where anyone can report something a student might have said and where the report automatically goes into the school’s surveillance database and case management system. And in my hypothetical, suppose someone anonymously reports via the website that a female student said to a group in her dorm, “I support Trump, and anyone who doesn’t is an imbecile.” Should the university keep this in a student file that will show up whenever an administrator wishes to inquire about the student for discipline, recommendations, or any other purpose?
Over 1300 colleges and universities use a system like this created and operated by a company named Maxient; there are other similar vendors, and some universities have created their own systems. Database systems have different modules for different activities. There are modules for standard disciplinary matters, for academic misconduct, for inappropriate sexual conduct, for behavior in the dorms-- and the module for reporting that a student has come across to others as somehow being biased. These systems not only keep permanent records on every aspect of student life, but also templates for virtually every communication to be sent to a student. They even have prompts as to when the next communication needs to be sent.
In my hypothetical, a bias response team receives notice of the anonymous entry and sees that the entry is about a political matter. If the bias response team suspends the student for being a Trump supporter, that would be clearly illegal, but let's take a milder, more typical case for our hypothetical: although the complaining party believes the statement was said in an offensive manner, the bias response team concludes it is protected by the First Amendment. They might just leave the report in the student’s file without doing anything more, meaning the student isn’t even aware that a negative report has been filed about her. She might not even be a Trump supporter at all, but she is not given a chance to correct any false information that was reported, perhaps even reported malicioiusly. Or, the bias response team asks her to come to their office for a voluntary meeting. They say the meeting is voluntary-- but if the student declines, they input a note saying the student refused to attend the requested meeting, perhaps typing in “X refused to cooperate.”
Virginia Tech said in a District Court hearing (p. 44) that its policy was to
Invite them to engage in a voluntary conversation. . . . If a student fails to respond to this message, or declines to meet with our office, no further action is taken and the student faces no consequences of any kind.
Members of the Virginia Tech bias response team have no disciplinary powers unless they might also hold positions in other parts of the student conduct office. Nonetheless, the negative information remains in the student's permanent record. Other administrators will be able to see the information if the student is accused of something else, e.g., drinking beer in a dorm, and this prior report will come up automatically when an administrator with power to punish her for that offense looks up her name.
Would such a bias response apparatus chill speech at the university? Speech First and three circuits say yes; Virginia Tech (Sands) and two circuits say no.1 The question presented by the petitioner is:
“Whether bias-response teams objectively chill students’ speech.”
Surveillance case management software is part of this question implicitly, since they are what bias response and other student conduct staff members rely upon. That could be clarified by restating the question as:
“Whether bias-response teams and the surveillance software upon which they rely objectively chill students’ speech.”
Speech First’s cert petition says,
“Precisely because speech codes are often struck down, universities have looked for subtler, more sophisticated ways to chill “offensive” speech.”
The Court must speak broadly, because universities are well known for trying to continue illegal activities by means the Court has not yet directly ruled on. We saw this after Brown v. Board of Education, and we are seeing it now after Students for Fair Admissions v. Harvard.2
The problem remains even if the bias response teams are gone Indeed, Virginia Tech has already eliminated its bias response teams. The cert petition says,
A month after the Fourth Circuit issued its mandate (and days after the University learned that Speech First was likely seeking certiorari), counsel for the University emailed counsel for Speech First, asserting that the University “has dissolved its Bias Intervention and Response Team.”
Even if Virginia Tech does not revive its bias response teams, students would still have to worry that their speech was being reported for university administrators to see and to remain in a student’s permanent files, without the knowledge of the student and without opportunity to correct wrong information, including information filed maliciously.
Does Surveillance Software Chill Speech?
The cert petition says,
The question presented is purely legal. It’s a question of Article III standing. And it turns on whether the University’s policy would chill the speech of a reasonable college student—an “objective” inquiry.
Let's turn to a more extreme hypothetical. A university hires student informants to wear hidden mics to record conversations with other students. Many states, including Virginia, are one-party consent states, in which this is legal. Stanford even offered monetary rewards for reports of wrongful speech as part of its Elimination of Harmful Language Initiative. Suppose the informants' reports are entered into a computer, where a voice-recognition app transcribes the conversations. An artificial intelligence app then searches for key words and deduces that the targeted student has a propensity to do things deemed inappropriate, rating the risk on a 1-9 scale from Bad to Good. Any administrator can see the rating and the conversations that generate it.
The system also uses a customized university version of ChatGPT, which is asked to write descriptions of students on campus who meet certain criteria, including the most inflammatory and unpopular things they have said. Assume further that the information is shared among a group of 60 other colleges and universities (not unusual since most of these systems involve cooperative efforts among schools and other entities nationwide) on the theory that AI works better with more data, and thus better systems will include speech from a large number of schools. The result of this wider data base, however, is that the biases of the schools themselves are being incorporated into AI’s own analyses, and thus a self-perpetuating feedback loop is created.
Some universities advise students of their rights under FERPA (the federal Family Educational Rights and Privacy Act) to review their files, but most schools do not provide such notice and thus in our extreme hypothetical, the student does not know that a file and rating might exist, or even that a surveillance and case management system is being used. Even if they know it exists, they would not know they themselves had a file and a social credit rating.
Would this offend the U.S. Constitution? All the university is doing is recording information legally collected, and doing some statistical analysis on it. Would it chill speech? We think so. But would the 4th Circuit think so? No--- it would be fine with them. The District Court said (from the cert petition):
Speech First lacked standing, the court reasoned, because the policy “do[es] not proscribe anything” and BIRT “lacks any authority to discipline or otherwise punish students.” App.107
And, indeed, in our hypothetical the university’s surveillance software does not proscribe anything. And we haven’t even said anything about a bias response team yet.
As for the 4th Circuit Court of Appeals (from the cert petition)
The majority found that Speech First lacked standing because Virginia Tech’s bias-response team has no “‘authority to discipline or otherwise punish students’” and so no “objectively reasonable student would self-censor to avoid encountering it.” App.19, 22.
In our hypothetical, the surveillance software has no authority to discipline or otherwise punish students, so the the majority thinks that any student who self-censored to avoid having his statements entered into the system would be objectively unreasonable.
In my opinion, it is the 4th Circuit, not the student, who is being objectively unreasonable.
August 17 update: I just heard that a doctor at Yale Medical School is actually proposing something like my hypothetical, but for Yale’s doctors:
"I was very upset in how bold I felt a lot of the anti-Black, racist statements I'd heard were since really starting as a medical student, frankly, and how common it was and how public it was behind hospital walls," she said.
Calhoun, a child psychiatry fellow at Yale Child Study Center at the Yale School of Medicine, said she was thinking about the effects of medical racism on patient outcomes when an idea popped into her head: what if medical professionals wore body cameras to provide documentation and create accountability?
"I felt like these people really feel comfortable saying this and it really is going unchecked."
That is, she seriously proposes that doctors at Yale University be required to wear body cameras to record their conversations in order to chill their speech!
Also in today’s news is a report that MIT is using grant money from a shady Chinese company to help it develop surveillance software to keep track of dissidents:
The Massachusetts Institute of Technology used funding from a twice-sanctioned Chinese company to advance components of facial recognition technology, which its Chinese benefactor has reportedly used to track and imprison Uyghurs. . . .
SenseTime's tech is part of a "vast, secret system" the Chinese use to "track and control" Uyghurs, the New York Times reported in 2019. The Trump administration that year blacklisted SenseTime for its role in the "repression, mass arbitrary detention and high-technology surveillance" of Uyghur Muslims in Xinjiang, preventing it from receiving technology or exporting it to the United States. Two years later, in 2021, the Biden administration banned U.S. investment in SenseTime.
Thus, we see that a pre-eminent American university is helping develope the capability to track their students’ political statements and have no qualms about someone using that technology for mass repression and genocide. Not all universities are as good at this as MIT, but they can no doubt buy surveillance software from SenseTime or other Chinese companies.
Those two news stories are just from one hour of browsing. We need to wake up. If your speech isn’t already chilled when you visit a university, it’s only because you are not aware of what they are doing yet.
How and Why Is Speech Being Chilled?
In his dissent, Judge Wilkinson said (as reported in the cert petition),
The “reality” is that the University “has constructed a complex apparatus for policing and reporting whatever administrators may deem ‘biased speech.’” App.39.
But why does that bother us?
The answer is obvious. It is because the system is all too easy to abuse. It has no stated purpose except to allow administrators to tell students to change what they say to other people. The decision as to what speech is deemed biased turns solely on what a single administrative office think is biased and they will have a propensity to inject their own political and social viewpoints into these highly subjective determinations.
And the biggest problem is that the administrators do not simply talk to the student-- they leave the complaint in the database. Indeed, the more outlandish the complaint, the more likely it is that they leave it there without the student ever knowing. Suppose someone asks you for all your credit card numbers and says he is going to put in a database that hundreds of people can access, but none of them can legally use the information to harm you. If you can really trust him, it would be objectively unreasonable to object to his request. But you can’t. He has no reason to ask for that information unless he intends to abuse it. Indeed, the very fact of his wanting the information is evidence of his ill intent.
To be sure, a university might have strict controls on who can see what is in the surveillance and case management software. It might, but it might not. How could a student know what negative and possibly false records are being created about him? In most cases, the universities do not tell the students, and even if they did, how could their assurances be verified? If a student somehow did learn that a malicious anonymous report was made about him and that the entire student services staff, and even the dorm employees, had access to the files, what would be his recourse? If he sued, what would his damages be?
Speech First v. Killeen, a case involving the University of Illinois, hints as to the situation, saying that the if the bias response team (BART)
determines the identity of the student who committed the bias-motivated incident, "it will record the details of the incident on the student's permanent record" and "will make this information available to others outside of the BART." Neily reports that one student's advisor told him "that he could see from the student's files that the student had met with someone from the BART."
Since 2023, university surveillance software has become a national scandal. An article in National Review says of Stanford University (my boldface),
The website says the process exists to address “situations involving real or perceived incidents” and even encourages students to report incidents that “may involve constitutionally protected speech.”
An anonymous source familiar with the use of Maxient at Stanford and elsewhere reports witnessing at another institution “a situation when the Title IX disciplinary person started to refer to past reports,” and when the student asked to see them, he “was told that information is confidential and he is not allowed to see it.” Asked about this, attorney Raul Jauregui, who works on Title IX cases, said, “Definitely this has happened to me during conversations with Title IX investigators and directors.”
Student-conduct officers can use the files, including complaints students do not even know about, as well as ones for which they are found not responsible, to establish patterns of behavior that can be used against them. Sydor noted that, “Stanford also states on the form that ‘This process was set-up to collect data’ which we know can be misinterpreted to create a pattern.” The anonymous source reported that, “If a student is accused of anything, the records are typically pulled up and then used in initial meetings with the student, in assembling a case for discipline and prosecution including Title IX but other disciplinary matters as well, and may often then even be brought to the hearing officers or hearing panels for the prosecutors to show this is a pattern and practice by the accused student.”
The information collected might also be reviewed when students apply for housing, resident-assistant positions, or other perks or jobs at their university, the anonymous source also suggested. Asked if he had ever seen something like this, Jauregui responded, “Yes. All the time. I’ve seen it go as far as the TIX office forcing students to resign from jobs that are at employers not owned by but certainly doing business with the school.”. . .
The anonymous source gave another example of how the way the software is used can prejudice this issue. This person had experiences in which “correspondence from the Title IX office and the relevant deputy dean was always being copied to Maxient, but when the student and his lawyer asked that their material also go to Maxient, they were told that only university people can make entries.” As a result, “accusatory information, whether correct or false, makes it into a student’s files but the exculpatory information often does not.”
Jauregui backed this up as well. Asked if it has happened to him, he again responded, “Yes. All the time. Either denied or they sort of forget for some reason to add it.”
So abuse is certainly a danger. How about the benefit of collecting information? Universities and students alike benefit from much of the information that universities collect in databases. To say that a university cannot keep files on students would be ludicrous. What is a grade transcript but a file on a student? And nobody would object to a record of which dorm the student lived in, or his phone number. The student’s disciplinary record is more sensitive, but clearly something the university should keep track of, as are physical and mental health records. Security and need-to-known access controls are important for all of this information, but nobody disputes the need for university databases.
What one must question, though is the benefit of keeping information on what a student says whether administrators have decided to add note saying it is “biased,” whatever that means. Is there any legitimate reason for the university to keep a file of a student's statements on political and social issues with which the university disagrees? The cert petition says of the 4th Circuit,
The majority praised the University’s bias-response team as “a way to educate [the] student body” about “harmful stereotypes,” “discriminatory tropes,” and “‘the role of tolerance in the campus community.’” App.27
That doesn’t fly. A database of critical remarks on protected speech and a bias response team to scold a student for something the student said is not a way to educate students, especially on the role of tolerance. It is a way to make them keep their mouths shut about controversial issues.
Consider another hypothetical. In 1965 a southern public university, formerly segregated, tells its students to “See Something? Say Something!” by putting anonymous notes in special boxes around campus if they see students of different races engaging in physical contact. The university acknowledges that such touching is perfectly legal and does not violate university rules. It says, though, that it wants to keep track of the behavior, with no disciplinary measures whatsoever, so as to establish a pattern if some future rules violation does occur. It hints that people who engage in this kind of physical contact are often the kind of people who break the law, so it’s important to keep track of them. It also says it wants to present data on the number of such incidents on campus. And it says that it wants to use bigot response teams to educate students about harmful stereotypes, discriminatory tropes, and the role of tolerance. Many students, unfortunately, are offended when they see an interracial couple holding hands. Such students should not be given offense unnecessarily. Their prejudices must be tolerated. They may suffer from false stereotypes and tropes, but they are sincere and they are pained to see such things. The university will not prohibit hand-holding, but it does need to educate the hand-holding couple in how to be more considerate of other people who might be offended.
“To educate the student body” is a feeble excuse. Why would a student’s education require files of supposed bias reports about that student that can contain anything any anonymous informant feels like submitting, even if the report (true or false) might permanently tar the student’s reputation? Remember, the files contain anything submitted, which might be as mild as “She said she voted for Trump and anyone who doesn’t is an imbecile,” hardly, we hope, something that puts the student in need of re-education. Of course, Biden supporters might disagree, but unless the university says explicitly that it is for Democrats only, it should not allow this kind of complaint into its records.
This does not even require appeal to the 1st Amendment. The student records privacy law, FERPA, says, (20 U.S. Code §1232g(b),
(1) No funds shall be made available under any applicable program to any educational agency or institution which has a policy or practice of permitting the release of education records . . . to any individual, agency, or organization, other than to the following—
(A) other school officials, including teachers within the educational institution or local educational agency, who have been determined by such agency or institution to have legitimate educational interests, including the educational interests of the child for whom consent would otherwise be required; . . .[Boldface added]
Let’s assume the student who was the subject of the report about her statement regarding Trump is reported a month later to have possibly been drinking wine in her dorm room. What “legitimate educational interest” does the dorm staff have in being able to see a prior anonymous report that this same student said she voted for Donald Trump and that anyone who didn’t is an imbecile? While FERPA itself (as opposed to the First Amendment) might not ban using software to collect such information, it would be a FERPA violation to allow anyone without a legitimate reason to have access to this educational record.
Once students know about the wide and nearly limitless scope of the kinds of reports that are going into these surveillance and case management systems, an “objectively reasonable student” (the test courts apply) would and should in fact fear speaking her mind, even in a dormitory “bull session” with her fellow students, all of whom might have differing viewpoints, lest it damage her future at the university and beyond.
The Remedy
What is to be done? Vague rules can always chill speech, especially when it is left to a single student services staff member, or even an entire office, to decide on their own that something a student said is somehow “biased.” It’s even worse when the entire surveillance and case management system is considered in its totality. Colleges have a lot of power and a lot of discretion. They may claim they support free speech, but what does that mean if they keep track of any opinions that someone else complains about and lets any one of a hundred people with power over the student look into the surveillance and case management software to see what complaints and other notations that have been about a student, including anonymously.
The remedies are straightforward.
First, forbid universities from maintaining records of a student’s speech unless it involves the breaking of university rules, or relates to a concern about the health of that student-- in which case it will be a securely private medical record.
Second, be sure strong firewalls exist and are periodically examined to assure that only people with a legitimate educational interest in specific areas have access to each module in the system. This is already required by law for medical records.
Third, forbid schools to poll the records of other schools to see if there are records about a specific student (at least one vendors actually enables this presently).
Fourth, notify a student if something is put in their file through a complaint form, and give him the right to submit correcting, clarifying, or exculpatory information.
Finally, expunge all student behavior records within some designated period of time after an individual student has graduated.
The point here is clarity and transparency. Virginia Tech is a state school and so may not prohibit speech protected by the First Amendment. Private universities could have more restrictive rules about speech although many states, such as California, prohibit schools from adopting limitations on speech that would not be permitted for a public college or university. Congress has said for all states, in 20 USC 1011a,
It is the sense of Congress that no student attending an institution of higher education on a full- or part-time basis should, on the basis of participation in protected speech or protected association, be excluded from participation in, be denied the benefits of, or be subjected to discrimination or official sanction under any education program, activity, or division of the institution directly or indirectly receiving financial assistance under this chapter, whether or not such program, activity, or division is sponsored or officially sanctioned by the institution.
Two Examples: Joshua Katz and Eric Rasmusen
What I have discussed above concerns student freedoms, especially their freedoms of speech and thinking. Allow me to shift to a parallel set of concerns which relate to restrictions on faculty speech and academic freedom.
One example is the case of Professor Joshua Katz of Princeton. Katz, a classics professor, published an op-ed in Quillette in 2020 in which he denounced an open letter signed by 350 faculty members and staff calling on Princeton to enact 48 demands such as hiring black faculty and divesting from businesses running prisons. As an op-ed in The Daily Princetonian says, President Eisgruber condemned Katz, showing his disdain for the Chicago Principles which Princeton had adopted:
“While free speech permits students and faculty to make arguments that are bold, provocative, or even offensive, we all have an obligation to exercise that right responsibly,” Eisgruber said in a statement to The Daily Princetonian. “Joshua Katz has failed to do so, and I object personally and strongly to his false description of a Princeton student group as a ‘local terrorist organization.’”
In 2018, Katz had been investigated for an affair with a student that was consensual but contrary to university rules, and he had been suspended temporarily without pay as a punishment. President Eisgruber had the case re-opened, and punished Katz again by firing him. The girl had not complained about Katz before, but in the meantime Katz had married a different former student and her attitude changed. She now complained that Katz had discouraged her from going in for mental counselling. Katz was fired for this new wrinkle on his old offense, Princeton said, but of course he was fired for publishing an op-ed.
In 2018, Katz had been investigated for an affair with a student that was consensual but contrary to university rules, and he had been suspended temporarily without pay as a punishment. President Eisgruber had the case re-opened, and punished Katz again by firing him. The female student had not complained about Katz before, but in the meantime Katz had married another former student, and the former student’s attitude apparently now had changed. She now complained that Katz had discouraged her from going in for mental counselling. Katz’s firing was obviously pretextual.
I myself am an example of how vindictive university officials like to keep records, often secretly. I was a member of the Indiana Association of Scholars, an Indiana University faculty group devoted to free inquiry. We embarrassed the Dean of the Law School, Lauren Robel, in some incident or other. I remember thinking maybe we should have asked her to fix the problem first rather than publicly embarrassing her. I do remember that the next incident, in which a School of Kinesiology (physical education) course taught by someone with an Arabic name devoted most of its syllabus to readings and sessions about how evil the nation of Israel was. The course, or at least the syllabus, was quietly fixed after our complaint. Later, I was elected to the faculty senate for a couple of terms. In the second term, Lauren Robel had become Provost, making her the chief academic officer of the Bloomington campus, the main unit of Indiana University. I frequently spoke and frequently voted against Administration policy, often in a minority of one or just a few. I also pointed out a couple of times where the chair, Provost Robel, was violating the rules of parliamentary procedure by shutting down debate without a vote. I think this annoyed her.
In 2019, complaints poured in about me because I’d quoted the statement “Geniuses are overwhelming male because they combined outlier high IQ with moderately low Agreeableness and moderately low Conscientiousness,” from an article titled, “Are Women Ruining Academia”. My quote was picked up by a 300,000-follower Twitter site called SheRatesDogs that specializes in stories about bad boyfriends. The day the fuss started, Lauren phoned me to say that she thought it only fair to say that she was issuing a public statement about me. In that statement, she said, among other things,
Professor Eric Rasmusen has, for many years, used his private social media accounts to disseminate his racist, sexist, and homophobic views. . . .
His expressed views are stunningly ignorant, more consistent with someone who lived in the 18th century than the 21st. Sometimes Professor Rasmusen explains his views as animated by his Christian faith, although Christ was neither a bigot nor did he use slurs; indeed, he counseled avoiding judgments. Rhetorically speaking, Professor Rasmusen has demonstrated no difficulty in casting the first, or the lethal, stone. His latest posts slurring women . . .
She also attributed to me a variety of political positions, some of which I do hold and some of which I don’t. I didn’t complaint about being called racist and so forth, but I did tell Lauren that she had gotten my beliefs wrong. She said that she could back them all up with evidence. She then opened an exhaustive investigation into whether I had violated any university rules, since there had never been any complaints in the past she could use. I retired (for other reasons) before it was over, but it appeared they were going to cut my pay 10% on a variety of wrongful pretexts such as complimenting an assistant professor on her dress and asking a Chinese student questions about his country. In my case, it was not surveillance software, but rather the custom-filing done for particularly disliked members of the university community. But this shows how such things work. For universities with thousands of students to surveil, a more mechanized process is needed— and such as is now provided by Maxient and companies like it.
Student surveillance and case management software is a serious problem. Once universities start collecting information on student protected speech, nothing can result but a climate of fear, with the turning of half the students into informants and the other half into fearful conformists.
Footnotes
See Speech First v. Fenves, 979 F.3d 319 (5th Cir. 2020), as revised (Oct. 30, 2020); Speech First v. Schlissel, 939 F.3d 756 (6th Cir. 2019); Speech First v. Cartwright, 32 F.4th 1110 (11th Cir. 2022) and compare with Speech First v. Sands, 69 F.4th 184 (4th Cir. 2023); Speech First v. Killeen, 968 F.3d 628 (7th Cir. 2020). I’ve put together a primary source link page at https://www.rasmusen.org/rasmapedia/index.php?title=SpeechFirst_University_Database_cases#The_4th_Circuit_Virginia_Tech_Case.2C_Speech_First_v._Sands.
See “Law School Org Hosts Conference On How To Keep Racially Discriminating In Admissions,” The Daily Caller (August 16, 2023).