Before 2024, you likely did not hear the terms “unalive” or “unaliving” outside of online spaces, like TikTok. That is, until Seattle’s Museum of Pop Culture used it in their very-real Nirvana: Taking Punk to the Masses exhibit to describe the passing of Kurt Cobain.
The 27 Club, one of the exhibit’s placards read. Kurt Cobain un-alived himself at 27, placing him in the company of other artists who passed at that same age under tragic circumstances…
In a world starkly divided by the digital and the “irl”, some new elements of human speech have become lost in translation. “Unaliving”, in particular, has sprung up as a substitute for the word suicide. Former Washington Post columnist Taylor Lorenz coined the term Algospeak, a combination of the words “algorithm” and “speak”, to capture the way online algorithms have begun to shape human language.
Algorithms serve as an essential function for social media companies. They adapt to users’ interactions online, promoting videos that align with individuals’ interests expressed through their likes, saves, and comments, to create a personalized viewing experience. Though, algorithms also play a large role in online safety. Particularly, in identifying and filtering out videos containing harmful content.
According to TikTok’s Community Guidelines, “harmful content” is defined as things like “violent and criminal behavior” and “harassment and bullying”. Read on to the “Mental and Behavioral Health” section and you will see this also includes: “We don’t allow content that shows, promotes, or provides instructions for suicide or self-harm.”
TikTok widely employs Artificial Intelligence (AI) to moderate its algorithms (Fang, 2023). To streamline efficacy, AI algorithms flag content with any explicit mention or depiction of suicide or self-harm and promptly removes it from the ever-swelling body of content created every day. While this guideline appears helpful at first glance, this process possesses a distinct lack of nuance that brings into question the harm of talking–and not talking–about suicide, and the potential repercussions of the terms rising to replace it.
So, what does the evidence say?
Algospeak is praised, primarily, for its prosocial effect. For one, it honors the adaptation of language as a natural linguistic phenomenon. Furthermore, algospeak’s development enables communities–largely, those marginalized–to commune online over topics of potentially shared experiences. Such discussions promote awareness, as well as reduce the stigmatization often surrounding these sensitive matters. Considering the unique ability for the internet to bridge geographic distance, this also means that individuals, even if isolated in a physical sense, can find comradery online. Thus, it helps connect individuals and provide the sense of support known to be protective factors (Bury, 2025).
With the very nature of algorithms being that they respond to users’ expressed interests, some evidence has emerged implicating algorithms as drivers of in-group bias as it discourages interactions with out-group members (Jawad et al., 2024). Whether intentional or not, this influence may promote mental health stigma amongst those who are not already familiar with the topic. Not only is potentially educating content stifled, but the broad identification of suicide and self-harm as “harmful content” in TikTok’s guidelines also creates a narrative that suicide is something to be avoided at all costs, that even the discussion of it is dangerous. It raises into question what we do as individuals when the media becomes an oligarchy, where the few in control also possess the decision to deem what is okay, and not okay, to talk about.
Thus, algospeak can also be viewed as a form of resistance. It creates an avenue for speakers to educate on the topic of suicide by using terms the algorithm would not recognize as “harmful”, enabling their videos to reach the audiences who may harbor these sentiments (Stano, 2023). This unique form of self-censorship preserves rapport and vulnerability, two fundamental therapeutic foundations, in the online hemisphere.
While algospeak may promote conversation, how that conversation is utilized is not innate. Some critique algospeak as a hollow attempt to bring attention to necessary conversations, like suicide, using the shock-value inherent to these topics to garner likes and views online. Novel research analyzing language and presumed purpose of the video found associations with algospeak usage to greater levels of harm in content, with lower levels of supportive intent, than videos with nonalgospeak (Schock et al., 2025).
Such prompts discussion around a sort of communicative capitalism, in which a “fantasy of participation” is created in the eye of the likes, comments, and views storm. Political theorist and political science professor Jodie Dean warns, “The proliferation, distribution, acceleration and intensification of communicative access and opportunity, far from enhancing democratic governance or resistance, results in precisely the opposite – the post-political formation of communicative capitalism.” When engagement with one’s content equates to success, creators are more inclined to use terms like “unaliving” to ensure their content reaches mass audiences and, thus, turn a profit, with less regard for the importance of the topic at hand. The phenomenon behind the mask of their algospeak terminology loses meaning over time, as any productive change that could be made on “corporate, institutional, and governmental” levels is stunted as one celebrates their impact by merely consuming this content (Dean, 2005).
From a clinical perspective, algospeak is a curious contrast to healthcare’s approach to suicide. It is a holistically debunked myth that explicit references to suicide increases suicidal ideation or the motivation to act on such thoughts. So much so, mental health clinicians are trained to ask with lucid specification when assessing suicide risk, as direct mentions correlate with reduced stigmatization and even seems to promote treatment-seeking behavior (Gould et al., 2005; Dazzi et al., 2014). One could potentially argue algospeak increases stigmatization by creating a “bad word” out of suicide.
Sitting with this evidence, the pros appear to outweigh the cons in the end. While algospeak does, seemingly, impede the clinical impetus of discussing suicide bluntly, one must keep in mind the context within which algospeak is employed. The internet is an undeniably crucial space in modern society where individuals across distances are able to find immediate community. When algorithms actively destroy any discourse regarding suicide for fear of allowing harmful content (an understandable, but still hindering precaution), people must find ways to preserve that connection–and more importantly, that conversation. Algospeak returns the power to the people to determine if their experiences are worthy of being heard, and any critique on algospeak, perhaps, should be redirected to the stonewall-esque algorithms that necessitate its existence in the first place.
On that note, however, algospeak should surely not make its ways into hospitals, let alone museums. Lest, of course, those museums specifically intend to enshrine pop culture, like punk bands, pseudoscientific behavioral patterns, and the quirky, deliberate methods humans have created to do the utmost human thing of all: Connect with one another.
~Isabella McIlvaine
Isabella McIlvaine is a rising 1st-year doctoral student in Clinical and School Psychology with experience working in both psychiatric research and healthcare settings. Her passions include creative writing, mental health, cooking, and animals!
References:
Bury, B. (2025). Decoding Algospeak: Unveiling the Patterns of Linguistic Evolution in the Digital Age.
Dazzi, T., Gribble, R., Wessely, S., & Fear, N. T. (2014). Does asking about suicide and related behaviours induce suicidal ideation? What is the evidence?. Psychological medicine, 44(16), 3361-3363.
Dean, Jodi. (2005). Communicative capitalism: Circulation and the foreclosure of politics. Cultural Politics (1)1, 51-74.
Fang, Serena. (2023). TikTok – Transform Entertainment with AI. Digital Data Design Institute at Harvard. https://aiinstitute.hbs.edu/platform-digit/submission/tiktok-transform-entertainment-with-ai/
Gould, M. S., Marrocco, F. A., Kleinman, M., Thomas, J. G., Mostkoff, K., Cote, J., & Davies, M. (2005). Evaluating iatrogenic risk of youth suicide screening programs: a randomized controlled trial. Journal of the American Medical Association (293)13, 1635-1643.
Jawad, M., Talreja, K.Bhutto, S. A., & Faizan, K. (2024). Investigating how AI Personalization Algorithms Influence Self-Perception, Group Identity, and Social Interactions Online. Review of Applied Management and Social Sciences, 7(4), 533-550. https://doi.org/10.47067/ramss.v7i4.397)
Schock, B., Kuraner, S. E., Westphal, M., & Bonanno, G. A. (2025). How “algospeak” is changing online discourse on suicide and self-harm: A pilot study. Traumatology. Advance online publication. https://doi.org/10.1037/trm0000581
Stano, Simona. (2023). “Linguistic guerrilla warfare 2.0: On the “forms” of online resistance.” Rivista Italiana di Filosofia del Linguaggio, 177-186. https://doi.org/10.4396/2022SFL13