Algorithmic indifference: The dearth of news recommendations on TikTok

Abstract

The role of recommendation systems in news consumption has been hotly contested. From one perspective, the combination of personalized recommendations and practically limitless content diminishes news consumption, as people turn to more entertaining fare. From another, algorithmic systems and social networks heighten incidental exposure, raising opportunities for news consumption regardless of explicit individual interest. In this work, we examine the potential for algorithmic exposure to news on TikTok, a massively popular social network built around short-form video. In the context of US-based news audiences, we examine the accounts TikTok recommends, the videos it shows new users, and its trending hashtags. We find almost no evidence of proactive news exposure on TikTok’s behalf. We also find that, while TikTok’s algorithms respond slightly to active signals of news interest from simulated users, that response does not lead to increased exposure to credible news content. These findings highlight a lack of algorithmic news distribution on TikTok.