Selected Papers of #AoIR2022: The 23rd Annual Conference of the Association of Internet Researchers Dublin, Ireland / 2-5 Nov 2022 Suggested Citation (APA): Rodrigues, B. and Peruzzo, A. (2022, November). Real Harassment, Virtual Robots Implications of Online Harassment Geared at Virtual Assistant Bia Paper presented at AoIR2022: The 23rd Annual Conference of the Association of Internet Researchers. Dublin, Ireland: AoIR. Retrieved from http://spir.aoir.org. REAL HARASSMENT, VIRTUAL ROBOTS? IMPLICATIONS OF ONLINE HARASSMENT GEARED AT VIRTUAL ASSISTANT BIA beatrys Cornell University André Peruzzo Universidade de São Paulo “For you, it can be a joke. For me, it was violent". This is one example of a new set of responses introduced by Bradesco Bank, the third-largest banking institution in Latin America, to their chatbot called BIA - Bradesco Inteligência Artificial (Bradesco Artificial Intelligence, acronym translated from Brazilian Portuguese). In April of 2021, Bradesco launched a new transmedia advertising campaign called "BIA's New Responses Against Harassment" (Bradesco, 2021), which featured a video that showcased BIA's less subservient responses to verbal attacks frequently sent to the virtual assistant. According to the bank, BIA received around 95,000 messages of offenses and sexual harassment (Propmark, 2021). The campaign had massive digital repercussions, especially the video on YouTube that had accumulated 168 million views, 17,508 comments, and over 44 thousand dislikes and 10 thousand likes (as of June 2021). Following worldwide trends in the gendering of AI assistants such as Amazon's Alexa, Microsoft's Cortana, and Apple's Siri, Bia is a common nickname for women called Bianca or Beatriz in Brazil. After its first initial advertising campaign, the chatbot was anthropomorphized and called by the pronoun "she," confirming its explicit feminine persona. However, feminist researchers have argued that the feminization of these assistants reenacts oppressive gender roles (Phan, 2019; Woods, 2018). The video was aligned with UNESCO's international initiative "Hey Update My Voice", a movement aiming at "pushing back against gender bias and sexual harassment of AIs" (heyupdatemyvoice.org, 2021). The initiative argued that these assistants are coded as feminine, on par with previous arguments. As Wajcman (2006) and Noble (2018) have postulated, technologies carry the signs of their designers and the culture in which they are developed so that they can reinforce gender, ethnic-racial, and class prejudices. This paper is an invitation to think about how the audience's negative reaction to the campaign illuminates the social and symbolic asymmetries in feminized work. One of our guiding questions in this research was: "Who is being dehumanized through BIA, and in turn, who becomes more human?". The comments help us to reflect on understanding what constitutes the human in artificial intelligence and how the harassment of this virtual chatbot relates to the harassment suffered by women in virtual settings. To critically analyze these texts, we have conducted a qualitative content analysis on 500 comments with the highest number of likes on the official Youtube video of the campaign, collected through Youtube Data Tools. By way of contextualization, it is worth noting that the repercussion of the ad on social networks is part of a broader context of polarization between the left and the right that has marked Brazilian political life, at least since 2013, from the protests held that year, through the articulated coup against President Dilma Rousseff, the rise of the extreme right represented by Bolsonaro and the recent election of Lula. In order to move away from technological determinism, we understand from Machado and Miskolci (2019) that these polarizations were not caused by social networks but made visible, catalyzed, and deepened by these same networks. These disputes have been especially characterized by the dissemination of moralizing guidelines in the sphere of sexuality and gender by various political-religious actors, based on what Miskolci defines as moral battles in the technical-mediatizing public sphere (MACHADO, MISKOLCI, 2019), in addition to other conservative agendas symbolized in the fight against corruption and a supposed communist threat. As communication scholars, we come from a tradition that sees "the semiotic construction of gender by different social technologies has concrete and material implications for people's lives" (De Lauretis, 1987). Also, we see advertising as a gendered technology that puts representations of femininity and masculinity in media circulation. Another productive concept is framing this ad as part of a broader trend of Outvertising. Mozdzenksi(2020, 2021) discusses how moral battles in the Brazilian advertising context are also expressed in the communication of brands, especially in a context in which advertising has been associated with social causes and guided issues of gender and sexuality. Outvertising is based on inclusive, tolerant, and pro-diversity discourses that denounce prejudice, stigma, discrimination, and violence against socially minority populations, even with significant tensions. We found intersecting but distinct themes that emerged from our thematic analysis of the comments. There is a recognition of the precariousness of female work and the harassment suffered by bank and call-center employees. Still, commenters are critical of the idea that robots can be victims of harassment. A large number of critical interactions and negative reactions demonstrate once again, even if metaphorically, mediated in the figure of a virtual assistant, the discomfort that any posture that contests sexism and violence may suffer on the internet. The commentators' resistance to change, the shameless reaffirmation of their harassment, and their constant irony emerge as expressions of “hegemonic masculinity” (Connell and Messerschmidt, 2005). A recurring theme in our sample could be explained as commenters feeling dehumanized while artificial intelligence is humanized. As one of our most liked comments said: "No friend if you messed with a robot, then you went too far, let's fight for Bia's human rights ah no, wait." This effort to place BIA within the boundaries of humanness enacted by the video, or as "someone" that can suffer harassment by other humans, may also work to conceal the lack of support that real call center women suffer every day within the bank. This argument seems to be directly linked with the technoliberalism ideology proposed by Atanasoski and Vora (2019). Looking for "technological enchantment" by automated objects, real labor is obscured by putting forth autonomous machines that seem to have a life of their own. Most of the commenters seem oblivious to the importance of symbolic harassment, at the same time, they support their objection by reclaiming the boundaries of humanness, especially the humanness that is denied women in assistant roles. As a sarcastic commentator said: "even artificial intelligence has more rights than me." This type of comment also shows how the audience understands exploitation. As Judy Wajcman famously proposed, "an emancipatory politics of technology requires more than hardware and software; it needs wetware - bodies, fluid, human agency" (2006). Through this, we show how the audience reflects on the contradictory understandings of the humans rendered invisible in technology and that changing the materiality/code of technology is not enough to affect systems of gendered oppression. As a final comment, the comments reflect on the mismatch between the issue of harassment raised by the campaign and the categories raised by the commentators. The campaign sought to use the BIA's new responses as an example of ways in which women can respond to offensive and unwelcome interactions and, in so doing, presented a less subservient female persona. The large number of critical interactions and negative reactions demonstrate once again, even if metaphorically, mediated in the figure of a virtual robot, the discomfort that any posture that contests machismo and violence suffers daily on the internet. The commentators' resistance to change, the shameless reaffirmation of their harassment, and their constant irony emerge as expressions of hegemonic masculinity. Similar to other moral panics and delegitimization of other examples of outvertising, the comments invest in efforts to silence the debate; create catastrophic and undesirable futures; and demoralize the "lacradores" - which are seen as crazy, without a sense of true justice. Bibliography #HeyUpdateMyVoice. (2021). Hey Update My Voice. https://heyupdatemyvoice.org/ Atanasoski, N., & Vora, K. (2019). Surrogate humanity: race, robots, and the politics of technological futures. Duke University Press. Bradesco. (2021, April 5). Novas respostas da BIA contra o assédio. YouTube. https://www.youtube.com/watch?v=-Ou6sCA1q1A&ab_channel=bradesco Connell, R. W., & Messerschmidt, J. W. (2005). Hegemonic masculinity: Rethinking the concept. Gender & society, 19(6), 829-859. De Lauretis, T. (1987). Technologies of Gender: Essays on Theory, Film, and Fiction (Theories of Representation and Difference). Indiana University Press. Machado, J., & Miskolci, R. (2019). Das Jornadas de junho à cruzada moral: o papel das redes sociais na polarização política brasileira. Sociologia & Antropologia, 9, 945- 970. Miskolci, R. (2021). Batalhas morais: política identitária na esfera pública técnico- midiatizadora. Autêntica Editora. Mozdzenski, L. (2021). Outvertising: a publicidade fora do armário. Editora Appris. Mozdzenski, L. (2020).“Mulher como garoto-propaganda do dia dos pais”: outvertising e as retóricas LGBTfóbicas na publicidade e no comentariado homotransfóbico brasileiro. Signos do Consumo, 12(2), 76-92. Noble, S. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press. Perez, C., & Peruzzo, A. L. S. (2018). Masculinidades em anúncio: a publicidade enquanto tecnologia de gênero. Tríade: Comunicação, Cultura e Mídia, 6(13). Phan, T. (2019). Amazon Echo and the Aesthetics of Whiteness. Catalyst: Feminism, Theory, Technoscience, 5(1), 1–38. https://doi.org/10.28968/cftt.v5i1.29586 Propmark. (2021, April 6). Bradesco mostra novas respostas da BIA contra assédio. Retrieved September 1, 2021, from https://propmark.com.br/anunciantes/bradesco- mostra-novas-respostas-da-bia-contra-assedio/ UNESCO. (2019). I’d blush if I could: closing gender divides in digital skills through education. https://unesdoc.unesco.org/ark:/48223/pf0000367416.page=1 Wajcman, J. (2006). El tecnofeminismo. Ediciones Cátedra. Woods, H. S. (2018). Asking more of Siri and Alexa: feminine persona in service of surveillance capitalism. Critical Studies in Media Communication, 35(4), 334–349. https://doi.org/10.1080/15295036.2018.1488082