The botization of science? Large-scale study of the presence and impact of Twitter bots in science dissemination

The rise of social media has brought new dynamics to the dissemination of scientific research, with Twitter (X) playing a significant role. This study focuses on the role of social bots—automated accounts designed to mimic human behavior and amplify content—in scientific communication. By analyzing over 3.7 million papers published between 2017 and 2021 and their 51 million Twitter mentions. Using a novel hybrid method that includes BotometerLite and specific activity parameters, with verification via a robustness check, it was found that 0.23% of accounts were bots. Despite their small numbers, these bots contributed to 4.72% of all mentions, indicating a significant presence, but with varied impact. Bots were particularly active in Mathematics, Physics, and Space Sciences, where they generated over 70% of tweets in some cases. Automated accounts disproportionately influence the visibility and perceived impact of research in these disciplines, which underscores the need for discipline-specific analysis when considering Twitter's role in scientific communication. This large-scale study highlights the potential for bots to skew altmetric indicators, misleading stakeholders about true engagement.