Bots and Binaries: On the Failure of Human Verification

In online tests of human verification, K-pop fans will often fail. A song looped on Spotify too many times, a track purchased and then re-purchased on the same music site, a YouTube video watched too-repetitively and without enough variation to bypass the platform's bot filter. In February 2020, YouTube removed millions of views from a newly released BTS music video as "part of standard precautions taken against bots," a move that directly affected the views accumulated by a fan-coordinated campaign to break the 24-hour debut record on the platform.1 In cases like these, the streaming and purchasing practices of K-pop fans code their engagement outside of so-called "human" digital behavior, a normative window drawn (and maintained) by platform algorithms and industry gatekeepers alike. Falling outside of its borders triggers the (mis)reading of fans not as human audiences but as bots: artificial figures whose streaming and purchasing habits exclude them from the category of genuine music consumption. As K-pop has risen in various Western music industries as a cultural force and chart competitor, the shadow of the bot, its accusation of non-belonging and illegitimacy, has only grown in English-language commentary on the topic.

Counterintuitively, I am less interested in refuting the techno-orientalist comparisons that Western commentators draw between K-pop fans and bot "hordes" or "armies" than I am in exploring the ways bothood (to borrow Taina Bucher's term) has served these fans in unexpected ways.2 I acknowledge, of course, that K-pop fans would be the first to argue for their own humanity and legitimacy, and rightly so in the face of its routine dismissal by the Western music industry. Yet, even as commentators cast these fans as a foreign and orientalized collective invading Western spaces of authenticity (the Billboard Charts, the Grammys), I can't help but speculate that industry discomfort with the threat of botlike fan collectives implicitly reveals the latters' force. Recasting the bot as a potentially powerful figure reframes the terms of the conversation away from techno-orientalist threat and toward an investigation of the way behaving as a bot — or being confused for one — can become an effective, flexible strategy for digital movement and grassroots organization online. 

As a member of these K-pop fan communities for nearly a decade, I have seen the way K-pop fans' adeptness with botlike strategy and behavior underwrites their digital power, the way it allows them to move with stunning efficiency both with and against the systems they choose to affect. When fans run hashtag promotions or spam campaigns, fundraise for social causes, or mass vote for awards, we can read these activities as functionally redeploying, or redirecting, the figure of the bot toward fans' own ends. And while the "goals" at stake in these campaigns often concern the financial and commercial success of their idol (and are therefore conservative in their orientation to capital and existing structures of power), they have also been demonstratively anti-hegemonic, anti-police, and anti-state in the service of marginalized populations — a product, arguably, of the way K-pop fans tend to identify with minority populations themselves.

These moments of resistance, and the more banal moments of community building, fandom in-humor, and everyday music consumption, underscore the ways that performing as a bot allows fans to move dynamically across digital platforms. To imagine fans as bots, or as botlike, is also to merge presumed opposites: on the one hand, the liveliness of fannish emotion, attachment, and excitement and, on the other, the flatness of the bot's mechanical efficiency and capacity for repetitive work. Merging the two asks how we might find unexpected abilities for organization and efficiency in the fan's lively feeling — and how, in turn, we can uncover deeply human ideas of passion and persistence in the bot's repetitive labor. To challenge the opposition of fans and bots, then, is also to interrogate the more basic binary of human and bots in the digital era: alive versus unalive, human against inhuman. In the context of a Western racial schema that associates the bot all-too-frequently with Asian identity and Asian digital presence, such binaries demand reevaluation.3 I follow, then, Michelle N. Huang's reversal of binary logic when she asserts not that Asians are robots but, rather, that "robots are Asian": a chiastic switch that uncovers the deeply human alterity that resides in the robotic.4 To reframe the (ro)bot in this way blurs the line between human and inhuman — a line that is always, Huang reminds us, "methodically engineered over and over again" — and begins, perhaps, to reverse engineer its racializing work.5

Mechanical Bodies

Techno-orientalism has colored the reception of K-pop artists in the West over the past decade. The figure of the bot looms especially large in English-language coverage of K-pop performance, hinging on the ease with which painstakingly synchronized choreographies slip, often unquestioned, into the orientalist figure of the robotic Asian body. U.S. media, for instance, approaches the K-pop idol with routine fascination and suspicion, often casting the idol body as empty of will and mechanical in its precision. Early in BTS' breakthrough into the U.S., just one year after their appearance at the Billboard Music Awards in 2017, The New Yorker published a piece to this effect titled, "Two Theories on How K-pop Made it to No. 1 in America." In it, Amanda Petrusich describes BTS as performers who "move with the kind of precision otherwise witnessed only at high-level cheerleading competitions," a vaguely neutral assessment — but one she follows with the admission that she "found it briefly disconcerting to see studied determination applied to something like club dancing."6 For Petrusich, K-pop performance treats the free expression of "club dancing" with a misplaced, studied gravity. Their dancing reads not as beautifully or impressively precise but as unnecessarily so: an uncomfortable "scruple" she must overcome to continue consuming BTS' media.7 As she goes on to describe BTS as "deeply, plainly deferential to their fans" by way of "a great deal of respectful bowing," Petrusich extends the discomfiting synchronicity of BTS' dancing bodies into their off-stage gestures and movements.8 The act of bowing, she explains, represents a staple of Korean culture, "and Chinese and Japanese and Vietnamese" cultures too, a mostly extraneous list of other Asian groups connected by the implicit excess of polysyndeton.9 Petrusich, in short, reads the K-pop idol's body as a broadly orientalist spectacle. She places the choreographed routine of their dancing alongside the Korean ("and Chinese and Japanese and Vietnamese") bow and, in so doing, understands them subtextually as one and the same. 

Petrusich's coverage continues the tradition arguably begun by John Seabrook's piece for The New Yorker, "Factory Girls," published in October 2012. As the title suggests, Seabrook leans heavily on a framing of the K-pop industry that emphasizes idol training, corporate oversight, and strategic marketing, casting these mechanisms as excessive when perceived through comparable systems in the US. Seabrook describes the vision of Lee Sooman, founder of SM Entertainment and the modern K-pop industry, as creating "a sophisticated system of artistic development" that "would make the star factory that Berry Gordy created at Motown look like a mom-and-pop operation."10 Citing Lee Sooman's background as an computer engineer, Seabrook emphasizes the way Lee merged an engineer's ethos with musical ambition to "create the blueprint for what became the K-pop idol assembly line."11 The "factory" and "assembly line" rhetoric underscores the implicit roboticism of the performing idols (many of whom, Seabrook speculates without direct evidence, likely received "plastic surgery") and the K-pop system as a whole.12 Published in the same year Psy's "Gangnam Style" achieved its viral momentum, Seabrook's piece, and others like it, had a palpable effect on the reception of K-pop in mainstream U.S. media — and continues to inform the reception of BTS and their fans half a decade later.

Where Seabrook belabored the factory-like roboticism of the K-pop idol scene, media commentators after BTS' 2017 breakthrough in the U.S. fixated on the figure of the bot, the robot's digital successor. As fans worked in concert with each other to build a grassroots promotional movement around BTS — and as this movement coincided with tangible, financial success on the U.S. Billboard charts — the accusation of chart "manipulation" from "botlike" fan tactics emerged from U.S. industry gatekeepers and media commentators. In one Buzzfeed News article, "Fans Are Spoofing Spotify with 'Fake Plays,' And That's A Problem For Music Charts," writer Blake Montgomery details the BTS fandom's campaign to place BTS' newest album, Love Yourself: Tear, at the top of the Billboard 200 chart via collective fundraising to buy premium Spotify accounts for fans who could not otherwise afford one.13 Framed as an exposé of "fraudulent" fan practices that placed the "gold standard" of the "widely respected" Billboard chart at risk, the article frames BTS fans as a "superfandom, millions strong across the globe" launching a "sophisticated campaign to make sure BTS reached No. 1."14 With language more suited to foreign military actors than an online fandom, the article inducts the collective of K-pop fans into the rhetoric of cyberthreat: a foreign mass, "millions strong," come to digitally game the system of a "widely respected" U.S. institution. Tellingly, the piece does not interrogate the long history of chart manipulation by corporate record labels within the U.S. itself — or Billboard's own symbiotic relationship to such practices. Nor does it mention the way BTS' chart success that year resulted largely from traditional physical album sales15 rather than high streaming numbers or clever Spotify strategies.16 In the article's view, only BTS fans' botlike and uniquely illegitimate methods could have accounted for BTS' success in the U.S.

Limits / Loopholes

My goal here isn't to deny that any gaming of the system occurs — neither is it to deny that K-pop fans do, in fact, maneuver in the gaps between official rules, prohibited practices, and technically possible behavior. Commentators' description of such behavior as savvy, manipulative, and even botlike can indeed accurately frame K-pop fan streaming and purchasing practices — although not in the way these commentators typically intend. To suggest as much is, I acknowledge, to move against the many years of fan labor correcting the misconception that any purchasing power stems from bots, or bot accounts, instead of human individuals organizing a legitimate form of grassroots support. Even the leader of BTS, Kim Namjoon, made an oblique reference to this running misconception of fans in an interview for the 2021 American Music Awards when he emphasized wanting to have a good time "with ARMY, with actual people—not AIs, not robots" at the show.17 The truth of K-pop fandoms' digital power, however, lies in a more complicated (and interesting) middle ground between the easy extremes of human or bot. 

While it's true that K-pop fans craft effective models of organizing across networks of real individuals and do not, for the most part, rely on automated bots to support their idols, their methods of organizing nonetheless follow a botlike logic to achieve their goals. In this sense, it would be more accurate to say that K-pop fans often move as bots across digital platforms without being, technically speaking, bots themselves. This distinction indexes the agency of fans as they navigate streaming platforms, music charts, award show criteria, and social media algorithms in undeniably savvy ways. To use the example of music charts: a popular fan strategy to improve digital chart performance entails buying a track multiple times across different music vendors to drive up sales — an echo of the bot's capacity for repetitive and wide-scale action. Because no current oversight can cross-reference (and therefore filter out or forbid) a download made on the U.S. platform, Spotify, with one made on the Korean platform, Soribada, a fan can download twice, support the artist in two (or more) venues, and increase streams accordingly. When fans do encounter a limit on purchases — such as the iTunes prohibition on buying any track more than once per account — they develop methods to creatively circumvent it. For iTunes, this might involve creating multiple accounts using temporary emails and buying the same track on each one. Here, the spambot's logic of maneuvering through the loopholes of the systems becomes entirely strategic: a way for fans to adapt in real time to obstacles or changes in platform oversight. 

It is important to qualify here, though, that fans develop these practices overwhelmingly within the parameters set out by institutions such as Billboard rather than attempting to hack, cheat, or recode the systems themselves. Per official Billboard chart rules, consumers may "buy a certain number of versions of songs or albums per week" to "contribute to the artist's weekly sales total or chart placement."18 The rules do not prohibit repetitive purchasing but rather limit how many times a single track or album may be bought to count toward the chart: until a change in 2022, for instance, the limit for the U.S. Billboard charts was four digital unit sales per transaction. Accordingly, a fan could purchase the same track or album four times (per approved vendor, per transaction, per week) and have all sales be counted as legitimate. Official chart rules, then, not only acknowledge but explicitly allow repeat purchases made primarily to improve an artist's chart placement — anticipating and legislating, so to speak, the extent of its own so-called "loopholes." The issue, then, lies more in the media framing of such activity as evidence for the illegitimacy of Korean artists' success in the U.S. than in the behavior of K-pop fans which, as Billboard themselves once put it, remains technically "above board" and within their own established parameters.19

"The Mysteries of the Machine"

Taina Bucher broadly defines "botness" as a mode of behavior or style of speech associated with algorithms and automated technology.20 Yet what has always struck me about her discussion of botness is not her definition, per se, but the illustrative quote she gives from Robinson Meyers's Atlantic piece on bots in the ensuing discussion: "an algorithm," Meyers writes, "does not need an audience."21 A Twitter bot generating random pieces of text is interesting precisely because it stems not from a clever person trying to impress but from the "network talking to itself about us": from human "inventions, speaking — somehow sublimely — of ourselves."22 Meyers describes the experience of witnessing it do so as "even a little voyeuristic."23 A bot that presents itself as a bot needs, after all, no audience and no obliging — it is a digital persona that "mutters to itself" apart from the rest, unconcerned with its position on the outside of human sociality.24 Admittedly, bots that attempt to pass as human do depend on, or perform for, an audience—but doing so speaks more to the humanness they try to emulate than to botness itself. The bot persona, insofar as it is botlike, encodes discordant and even asocial behaviors. I find this independence from audience intriguing for the way it expands the parameters of speech: the bot can speak against social expectation; it can transgress conversational etiquette — it need not even make sense.

The bot, then, offers a performative persona that can express what the "human" cannot or will not say, both on the level of form (the keysmash, the glitch, pure gibberish) and content (the unpopular opinion, the censored critique). Where the natural voice and human-coded speech reach the limits of its expression, the bot offers us new and additional possibilities for circumvention and escape. It offers, too, a certain inscrutability (evoking what Bucher calls "the mysteries of the machine") that all bots have, a technological aura that deflects understanding and impedes observation.25 It is this opacity — this mystique, this independence from social speech — that I see as a critical part of the bot's political and minoritarian potential.  

For Bucher, the "mysteries of the machine" are contained in the unpredictable and emergent behavior we often associate with digital programming: their performance(s) of non-human autonomy that crafts the power of the bot in our collective imagination. A bot program coded to randomize song lyrics might one day hit on an unexpectedly poetic combination. A chatbot designed to help users with a corporate website might accidentally insult or offend them. While these autonomous moments do not negate the fact of a bot's human-authored programming, it is the uncertain discrepancy between the two — the unresolved tension between human direction or artificial emergence as the origin of digital behavior and the openness of the bot to misreadings along these lines — that creates such a productive opacity around the figure of the bot. To strategically take up the figure of the bot, then, means to take on the ability to "pass": as an algorithm or program when one is human, and as human, when one is not. To cross back and forth along these lines is to move across the ontological and definitional boundaries of the human itself and becomes a profound opportunity for disguise.

Fan protest activity in support of the Black Lives Matter protests of 2020 leveraged this opacity well. As fans mass-tweeted against police brutality (boosting relevant protest hashtags and alternately hijacking white-supremacist hashtags and surveillance apps with a flood of K-pop-themed spam and images), fans coordinated multiple anonymous spam campaigns that deliberately confused or buried information on the opposing side.26 This political work exists not separately from the daily activities of K-pop fandom and their music-chart aspirations but on the same spectrum of digital savvy. After all, K-pop fandoms, from their position outside of the Western music industry, have long honed their ability to infiltrate, move, and affect its systems of power. They've learned to do so specifically through those loopholes, gaps, and technically not-unallowed avenues that institutions have left open to influence. Maneuvering along and within these gaps is as much a botlike strategy as it is a minoritarian one. 

As fans use loopholes and allowances to gain Western success for various groups, however, it is worth noting that they pursue the very routes designed to facilitate corporate record companies' exploitation of the same. In one sense, this means that fan agency at some level does collapse into the logic of capital, appropriated to the same demand for higher sales and profit. In another, however, it is worth emphasizing that before fans launched their initiatives in the U.S., such chart tactics had been largely the domain of corporate labels. The practice of "bundling," for instance, historically allowed record labels to tie an automatic or "included" album to the purchase of artist merchandise or concert tickets that would count toward the official Billboard charts. Bundling existed as a practice for decades until Billboard (only recently) banned it in 2020, with the admission that the "strategy [of bundling] has been employed by artists and labels to try and boost album sales," particularly in a digital era where such sales "are worth considerably more than streams on the charts."27 The intervention of K-pop fans, then, is not in the chart gamification itself but rather in the way they transferred this gamification into the hands of listeners and consumers — and did so on a scale, and with a level of organization, that produced culturally visible and paradigm-shifting results. Critically, their grassroots success did not benefit US record companies directly (at least, not at first) but redistributed much of the financial profit outside of the U.S. to entertainment companies in South Korea. It would be fair to say, then, that while K-pop fans do not undermine capital at large, they do subvert a specifically US and Western hegemony. Their botlike tactics of music consumption and advocacy decenter US structures of cultural power (and Anglophone expectation) to make politically meaningful room for non-English speaking celebrities and South Korean performance.

Humanizing the Bot

I will end with ruminations on the parallel between bots and political work — namely, Jennifer Rhee's reminder that our imaginaries of (ro)bots do less to secure some objective distinction between human and artificial than they do to construct an imaginary of "the human" that is "valorized and normativized."28 Against this vision of the human, those deemed (ro)botic fall into categories of "unknowability, unfamiliarity, or illegibility" — distinctly minoritarian categories of being that make possible their exclusion, erasure, and dehumanization.29 It is no coincidence that the bot's failures of "human verification" render it alien and even threatening: a figure of digital non-belonging.

Yet the activity of K-pop fans reminds us that the bot's failure to belong is precisely what opens it to reclamation, movement, and play. Even banal moments of community in-humor and fandom conversation speak to this possibility. The key-smashed gibberish posted by fans in response to a picture of their favorite idol ("SwhvsyWHAT /;i52#*{.#|*"), the joking replies that "this is not helping the 'kpop fans r bots' agenda" by fellow fans: such moments cast the inhuman, spam-like speech of bots not as nonsense but as a community-specific way of conveying overwhelmed, hyperbolic, or excessive feeling.30 The same might be said of fan accounts that repetitively retweet or repost an idol's images — sometimes by the hour, on the hour, as a spambot would — for the way they operate as a community-recognized form of dedication and promotion. To my mind, these routine behaviors create a provocative merging of fannish liveliness with the bot's automated inhumanity. It triggers a re-reading of the bot, one that finds devotion in its repetition, affective intensity in its gibberish, and love in its labor. To locate the human in the bot in this way, then, is to recode the bot's nonbelonging as the ideal register for nonstandard feeling: those affects too excessive, too femme, too raced, or too queer for so-called "human" behavior. The bot becomes a vehicle for nonconformist expression: the venue by which we perform a different kind of humanity.

In the fan's strategic slippage between person and bot — human and inhuman — we begin, perhaps, to see a path through the wider conflation of minority groups with dehumanizing languages of swarms, bots, or hordes. To turn from rejecting the label of the bot (derogatory, threatening) to embracing its implicit challenge to the human (valorized, normalized) is to re-code the binary itself, its fundamental demand for a degraded term. In doing so, we might make way for a newly intermediate figure of the digital era: the human-as-bot, a liminal and minoritarian possibility.

---

Andrea Acosta (@a_priyd) is a current PhD candidate in English at UCLA and incoming Assistant Professor of Digital Media at Pitzer College. Her research focus is on the intersection of digital media and critical race studies.

---

References