This was originally published on January 11, 2024 on my Substack newsletter. I’m republishing it here to get this post on my platform. (Never trust platforms you don’t own.) If you would like to support the work I do (I don’t get research funds now that I’m not faculty), I’m running a sale on newsletter subscriptions: a year for $24. Offer is good until 23 February 2024.
We are barely two weeks into 2024 and the mainstream Discourse has been exceptionally occupied with questions of academic honesty. After right-wing activists remixed GamerGate’s disingenuous concern with “ethics in game journalism” into a concern with potential academic dishonesty in Harvard President Claudine Gay’s dissertation, billionaire Bill Ackman had a Twitter meltdown over Business Insider’s finding that his wife, MIT Media Lab faculty Neri Oxman, plagiarized much more extensively in her own dissertation. Ackman threatened to run the work of everyone at MIT through plagiarism-detection software in retribution.
At the same time, Dr. Charlotte Lydia Riley noted on Twitter/X that industry-leading plagiarism detection software “Turnitin has carefully noted that it thinks a student has submitted an essay that is 11 per cent plagiarised. It has come to this figure by highlighting every time they have used the word “this”.” Treating the repetition of a commonly-used part of speech (a demonstrative pronoun if you want to get technical) as evidence of academic dishonesty, TurnItIn is an academic form of stop-and-frisk, criminalizing those on the low end of the student-teacher power differential just as stop-and-frisk policing criminalizes Black and brown people. TurnItIn flagged Riley’s student not so much for what they did – using a pronoun repeatedly is not a violation of academic dishonesty – but for who they are, i.e., a student. Likewise, Ackman is trying to use plagiarism accusations and codes of academic honesty as weapons to criminalize academics who either are racial/sexual/gender minorities and/or work in fields that have become increasingly inclusive of both members of these groups and their intellectual traditions. The reason his wife must be innocent of plagiarism is that she doesn’t fit the profile he has of an academic “criminal.”
Beyond Ackman’s spectacular and risible performance on social media, what’s interesting in all this is how intellectual property is being used to police race, gender, sexual, class, etc. status differentials through the criminalization of knowledge work and knowledge workers. As Anjali Vats writes in The Color of Creatorship, “Intellectual property law is also a ‘racial project,’ that reproduces particular racial orders, in which people of color are coded as lacking the capacity to create” (3), i.e., to be legitimate owners of intellectual property. Studying this racial differential in U.S. intellectual property law up through Obama-era postracialism, the most recent structure of “intellectual property citizenship” articulated in Vats’s book is one characterized by the “persistent nostalgia for histories of white male making, alongside foreign policy invocations of hyperracial infrignement, which blamed other countries for stealing America’s most valuable commodities” (115) – think fears over bootleg Chinese DVDs of Hollywood films, and so on. Much in the same way that Bush 43 framed the “terrorist” figure as an inherently foreign, non-Western threat to the purportedly American way of life in order to prop up the myth of pinkwashed, post-feminist American postracialism, this structure of intellectual property citizenship criminalizes foreign actors. 2024’s plagiarism-palooza is a way to criminalize the very sorts of educated, bourgeois and/or upwardly mobile figures like Claudine Gay or the U.S.’s increasingly Black and brown population of undergraduates, whose success otherwise served as evidence of the fact of American postracialism.
As such, plagiarism-palooza represents another way that the power relationship known as “securitization” and the tools of the security state is applied internally to police groups who otherwise benefit from postracial, postfeminist, homonatiationalist “inclusion.” As Tara García Mathewson writes in The Markup, language models from the EdTech sector make such copyright infringement appear to be increasingly easy. In the early 2000s when I started teaching undergrad classes, I had to be able to identify potentially plagiarized content based on my own humanly limited judgment and then Google suspect phrases myself to see if they came back with any hits to, for example, SparkNotes or Wikipedia (or my friends’ published academic research). That ability came from my years of study and training in a particular academic discipline. With TurnItIn and similar tools, any non-expert (like Bill Ackman) can have a go. While these tools are often thought to detect plagiarism, Matthewson clarifies that that’s not what they actually do. TurnItIn says it offers a “Similarity Report” that identifies the degree to which a submitted piece of student writing is similar to other texts in their database. In other words, TurnItIn creates what is more or less a criminal profile for a paper much in the same way that recidivism algorithms create data profiles of persons experiencing incarceration. These profiles aren’t evidence of factual transgressions of the law, but suggestions for the likelihood of possible transgression (based on racist and cisheterosexist bias built into the technology). As Foucault defines it, securitization is a mode of governance focused on the prevention of possible risk, often through practices of criminalization. By creating criminal profiles of academic papers, plaigarism-palooza applies techniques and practices of securitization forged in mass incarceration and the war on terror to domestic postracial/post-feminist/homonationalist elites and upwardly mobile groups.
This attempt to use a creative work’s profile as evidence of infringement parallels developments in the music industry, where there is an increasingly common tendency to shift the measure of infringement from exact copying to a sufficient degree of similarity in profile. Traditionally, copyright infringement has been understood as exact copying; hence Vanilla Ice’s infamous claim that “Ice, Ice Baby” didn’t copy Queen & David Bowie’s “Under Pressure” because “ding-ding-ding-ding-dinga-ding-ding” is how “theirs goes” and “ding-ding-ding-dinga-ding-ding-tss-ding, ding-ding-ding-ding-dinga-ding-ding” is how “ours goes.” Pointing out the extra “tss-ding,” Ice highlights the differences between the two bass riffs in order to discourage the perception that his song directly samples or copies “Under Pressure.” However, the plaintiffs in the most recent suit against Ed Sheeran treat copyright as a fuzzier matter that extends beyond exact copying. Though plaintiff’s argument in the Sheeran lawsuit claimed his single “Thinking Out Loud” has ““striking similarities” and shares “overt common elements”” with Marvin Gaye’s “Let’s Get It On,” the jury decided in Sheeran’s favor. Similarly, though there is a consensus among music critics that when it comes to the relationship between Beyonce’s “Break My Soul” and Robin S’s original “Show Me Love” “there is no debate at all that the original version and Beyonce’s new song sound nothing alike,” Beyonce and her team decided to preemptively credit Robin S because the songs share a similar profile. As these two examples suggest, there’s a move among some in the music industry to treat non-exact resemblance as copyright infringement. From this perspective, “Ice, Ice Baby”‘s “tss-ding” wouldn’t save Vanilla Ice from a copyright lawsuit because the overall profile of his riff–its rhythm, timbre, meter, tempo, and pitch–is aligned with the overall profile of the riff in “Under Pressure.” As many critics of the lawsuit against Sheeran note, this move would put all songwriters at the mercy of current musical asset owners, who would be owed royalties and credit for mere similarities to songs whose copyright they own.
AI excels at exactly this sort of profiling or perception of fuzzy similarity. Take, for example, ChatGPT, which is (in)famous for getting matters of fact nearly right. For example, when it first came out, I asked it to write a biography of philosopher Robin James. Although I received my Ph.D. In philosophy from DePaul University, Chat GPT flip-flopped between Penn State and Vanderbilt as my doctoral alma maters. The philosophy departments and doctoral programs at all three of these universities share a methodological orientation, which is evinced by the composition of my former department at UNC Charlotte: there was me, of course, and four colleagues with Vanderbilt Ph.D.s two of whom who used to teach at Penn State. My actual biography and my ChatGPT-generated biographies have the same overall profile, even though the relationship among the exact details is fuzzy. As this example illustrates, ChatGPT reads in the same way that the plaintiffs in the Sheeran case–Structured Asset Sales (SAS), who acquired a portion of Gaye’s co-writer Ed Townsend’s estate–want jurors to hear music: no details, just vibes. AI’s ability to perceive the very sort of fuzzy similarities that corporations like SAS want to establish as copyrightable threatens a future where companies like SAS use a bot like YouTube’s copyright-sniffing algorithm on steroids to shuffle through everything released on Spotify or Bandcamp and serve musicians large and small with copyright infringement lawsuits…just as Ackman threatened to check all the published work of faculty at MIT and the Ivy League.
Ackman’s anti-plagiarism crusade isn’t just an attempt by a right-wing billionaire to wage his front on the 21st century culture wars; it’s part of a broader mode of governance that uses the copyright vibe police to both criminalize the knowledge workers otherwise often included in postracial/postfeminist/homonationalist privilege and to steal their property. Claudine Gay is pilloried for a few stock-ish phrases in the acknowledgements section of her dissertation (the “thank you” section where no actual scholarly work is done), but Spotify is getting away with de-monetizing any track that gets less than a thousand streams a year. In this respect, Ackman is less a right-wing extremist and more a quintessential example of an increasingly common form of power which uses automated copyright infringement profiles to criminalize knowledge workers and creators and re-create patriarchal racial capitalist status hierarchies in terms that don’t explicitly reference identity.