The problem with porn filters

A mistake at Arizona State University

Early last week, students at Arizona State University discovered that a web filter has been implemented to block access to “adult” websites. Rachel Leingang, a local journalist, reached out to ASU about the filter, and ASU immediately pulled it, saying through a spokesperson:

The WiFi is back on tonight to unrestricted access. There was an inadvertent change applied to the WiFi network, which would normally only be applied on the student portion of the network in the summer, when we have a large number of underaged individuals and families on campus with WiFi access. Typically we only restrict access to sites that are known to be sources of malicious software.

As a public university, ASU is obligated to uphold the First Amendment, and as one of FIRE’s “green light” schools, ASU maintains no policies that threaten freedom of expression. Further, ASU has taken the additional step of committing to free speech by adopting a statement of principles modeled after the “Chicago Statement.” FIRE is pleased that ASU was quick to disable the filter when it was brought to their attention.

Nevertheless, that such a filter might be in place during the summer on the “student portion of the network” is troubling. The presence of “underaged individuals” on campus would not be a valid justification to shield adult students, who are also on campus over the summer, from protected speech. Freedom of expression and academic freedom aren’t seasonal.

The issue at ASU calls to mind the student-led efforts at the University of Notre Dame to get the university to implement a porn filter, which led to copycat efforts at Harvard University, Princeton University, and the University of Pennsylvania. While these efforts don’t appear to be bearing any fruit, a porn filter would violate each school’s promises of free expression to its students.

Since this is a trend that seems to be gaining steam, we’d like to discuss the problems and implications of such a web filter in a university environment.

Obscene or not obscene: that is the question

A porn filter at a public university is difficult to square with a public university’s obligations to the First Amendment.

Making the distinction between obscenity, a category of speech unprotected by the First Amendment, and the large majority of content that can be considered “pornographic,” which is protected, is notoriously difficult for humans, let alone web filters. Recall Supreme Court Justice Potter Stewart’s famous statement regarding the difficulty of defining obscenity from the 1964 case Jacobellis v. Ohio:

I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description [“hard-core pornography”], and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.

Web filters require discrete rules and definitions and judge content on a binary basis. If humans cannot effectively define the rules, the result will be an over- or under-inclusive filter. The legal test for what content is obscene, from the 1973 Supreme Court case Miller v. California, requires, in part, an evaluation of whether a piece of work lacks “serious literary, artistic, political, or scientific value.” This type of evaluation obviously cannot be made by an algorithm.

A net both too wide and too full of holes

Obscenity has a narrow definition and a complex test because the courts have held that the First Amendment obliges that any government restriction on speech be narrowly tailored, meaning that it must not incidentally restrict large amounts of protected speech. Government restrictions on speech are often struck down because of their over-inclusive definitions of what is prohibited.

However, because web filters are typically designed for either non-academic workplaces, or to keep children from accessing inappropriate material, they are not designed to block only the obscene, but rather the much wider and more ambiguous category of “adult content,” as was the case at ASU. Where business productivity and the protection of children are concerned, web filters are designed with a “better safe than sorry” philosophy, where over-inclusiveness is arguably a virtue rather than a vice. This results in the dragnet catching large amounts of content that is not only not obscene, but also not even “adult content.” Such over-inclusiveness is anything but narrowly tailored.

This is especially true because most common filters block adult content not based on the content itself, but the website on which the content is hosted. While this makes intuitive sense for websites centered around adult or pornagraphic content, this poses a particular problem when applied to massive websites such as Reddit and Twitter, which allow adult content, but are not primarily centered on it. If they are blocked by the filter, an enormous amount of non-adult content would become unavailable to users. If those websites are not blocked, they’re an easy way to evade the filter.

A blow to academic freedom

Even if one were to grant that a porn filter would make sense in another context, a university is likely the place where it makes least sense, due to the role of the university in researching and generating knowledge. This role is protected by the principle of academic freedom. Bans on access to online pornography effectively hobble a wide galaxy of research not only on pornography, but on sexual development, relationships, and other sociological or psychological topics. Academics may sometimes study that which is of prurient interest to their subjects.

Even if exemptions can be made for research work, it’s not hard to imagine how a system where students and professors have to apply for dispensation from the web filtering authorities would create needless embarrassment, stigmatization, or opportunities for censorship. Professors and students must be free to pursue research on the important questions of the day, and they should not need to leave campus, circumvent a filter, or apply for permission to do so.

Too easy to get around

As we’ve written before, web filters are generally ineffective at achieving their stated goals because of the ease of circumvention. Since universities can only filter their own networks, evading the filter can be as easy as going off campus, browsing on your mobile phone’s network or through tethering to your computer; or by using Tor or any number of virtual private networks (VPNs) or anonymizers to mask your traffic from the university.

The inconvenience burden is also more heavily borne by academics and researchers. Most porn traffic now comes from mobile devices, making blocks less of a problem for those who are consuming it out of “prurient interest” and more of a problem for those who need access to pornographic sites for academic research, which is significantly more difficult to do on a cellphone than a computer. (To be clear, the fact that such blocks might be easier to get around is not a reason to allow the blocks anyway. If the government were to ban a newspaper from printing paper copies and attempt to justify its action by saying that people could just go and read it online, neither the law nor an ordinary person would be likely to accept that reasoning.)

Additionally, there are literally millions of pornographic websites. Logistically, it’s impossible that a web filter will be able to block them all, at the rate that new websites pop up. While the filter may be successful at blocking the “mainstream” sites, it’s entirely possible it would simply push traffic to more niche websites.

What’s next?

For all the reasons stated above, public universities, and private universities that claim to value academic freedom and free expression, should not implement filters on adult content.

FIRE is currently researching the prevalence of porn filters on college campuses. If your college or university has a filter blocking adult content or pornography, FIRE would like to hear about it. Let us know through our case submission page.