Framing Gender Through Words: Examining Stereotypical Language in AI-Generated News Headlines on TikTok
Abstract
The increased integration of artificial intelligence in the development of digital news has had a significant impact on how information is concluded and consumed on social media platforms. TikTok, in particular, has become an important platform for the spread of information in the form of news, for which short videos accompany AI-generated headlines, which have a strong impact on the interpretation of the audience. This study examines the frame of gender in cases of AI-generated news headlines in TikTok and analyses whether the automated texts replicate stereotypical gender depictions. Employing a qualitative content analysis of 100 TikTok videos published from 2023-24, the study examines the insistent stability of lexic patterns in the headlines with the subjects (male vs. female) in particular manifesting certain aspects like authority, achievement, emotional framing, appearance, and relational identity. The results show the presence of clear and systematic gender asymmetries: Headlines of men apparently focus mainly on competence, agency, and professional accomplishment, while those of women more often foreground emotional states, physical appearance, and familial or relational roles. These patterns suggest AIs will not only reflect age-old gender issues in traditional news media, but they can actually exacerbate these issues in algorithmically led but engagement-optimised institutions. The study concludes that the unchecked acceptance of AI-generated language in the social media space for news content runs the risk of perpetuating gendered hierarchies as well as highlighting the need for more critical oversight in AI-assisted news content.
References
- Macharia, S. (2022). Who makes the news? Global Media Monitoring Project 2022 report. World Association for Christian Communication.
- Kotek, H., Dockum, R., & Sun, D. (2023, November). Gender bias and stereotypes in large language models. In Proceedings of the ACM collective intelligence conference (pp. 12-24). https: //dl.acm.org/doi/abs/10.1145/3582269.3615599
- Bender, E. M., Gebru, T., McMillan-Major, A., & Mitchell, M. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the ACM Conference on Fairness, Accountability, and Transparency, 610–623. https: //dl.acm.org/doi/abs/10.1145/3442188.3445922
- Sheng, E., Chang, K. W., Natarajan, P., & Peng, N. (2021). Societal biases in language generation: Progress and challenges. Proceedings of the Association for Computational Linguistics, 4275–4293. https: //aclanthology.org/2021.acl-long.330/
- Sun, T., Gaut, A., Tang, S., Huang, Y., ElSherief, M., Zhao, J., Mirza, D., Belding, E., Chang, K.W., and Wang, W.Y. (2019). Mitigating gender bias in natural language processing. Proceedings of the Association for Computational Linguistics, 1630–1640. https: //arxiv.org/abs/1906.08976
- Dawson, S. (2024). You can’t say that on TikTok: cxnsxrshxp, algorithmic (in) visibility, and the threat of representation (Doctoral dissertation, University of British Columbia). https: //open.library.ubc.ca/soa/cIRcle/collections/ubctheses/24/items/1.0443761
- Gillespie, T. (2022). Custodians of the internet: Platforms, content moderation, and the hidden decisions that shape social media. Yale University Press. https: //books.google.com/books?hl=en& lr=& id=cOJgDwAAQBAJ& oi=fnd& pg=PA1& dq=Gillespie, +T.+(2022).+Custodians+of+the+internet: +Platforms, +content+moderation, +and+the+hidden+decisions+that+shape+social+media.+Yale+University+Press.& ots=PiRLAYQQm0& sig=rmVfPtqQTFSDWdvjE_VI5etc-sw
- TikTok videos analyzed (2024). AI-generated news headlines retrieved from TikTok public accounts using keyword-based sampling (January–March 2024). https: //support.tiktok.com/en/using-tiktok/creating-videos/ai-generated-content
- Binns, R. (2018). Fairness in machine learning: Lessons from political philosophy. Proceedings of the Conference on Fairness, Accountability and Transparency, 149–159. https: //proceedings.mlr.press/v81/binns18a.html
- Zhao, J., Wang, T., Yatskar, M., Ordonez, V., & Chang, K. W. (2017). Men also like shopping: Reducing gender bias amplification using corpus-level constraints. Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2979–2989. https: //arxiv.org/abs/1707.09457
- Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press. https: //www.degruyterbrill.com/document/doi/10.18574/nyu/9781479833641.001.0001/html
- Van Dijck, J., Poell, T., & de Waal, M. (2023). The platform society: Public values in a connective world (2nd ed.). Oxford University Press.