Bot comments Report January 2025
The dataset of toxic bot comments detected in January 2025, spanning from 01 January to January 30, included over 600 English messages (n=629). The average tocixity score of the bot content was quite low, at 0.15. However, there is a notable peak on 21 January 2025. Click here for a list of possible events that could have influenced the peak. Only about two messages were very toxic, with a toxicity score higher than 0.8.
Toxicity frequently involved: Profanity (20%), Politics, Racism and/or Ridicule
Threatening content: About 4% of messages were classified in this category
Disinformation content: About 2% of messages were classified in this category
Most frequent category combination: Politics + Religion (see category explanations here)
Despite the low average toxicity, several messages contained problematic and harmful language, featuring toxic keywords such as 51st state, coward, drumpf, german, commie, and shithole.
The word cloud below highlights the 50 most frequently used toxic keywords in January 2025. The size of each word reflects its frequency relative to other terms. Notably, trump was the most frequently used keyword, followed by god, doll, chinese, and zionist.
The table below highlights the different toxic categories detected within the January 2025 dataset, along with their respective frequencies (represented by black dots and further explained below). In this month, the categories Ridicule, Contempt, Racism, Sexism, Politics, Religion and Threat were significantly more prominent than the others. However, content categorized as racist had a notably higher toxicity score of 0.30 than the other categories (on a scale from 0 to 1, where 1 represents the highest level of toxicity).
The bar chart below highlights that the most prominent categories with the highest toxicity scores in January 2025’s dataset of toxic bot comments were Profanity, Ridicule, followed by Sexism and Politics. The majority of the toxic bot content fell within the Politics category, indicating that much of the bot activity was aimed at influencing political discussions (see the explanation of the different categories here).
The keyword network below illustrates the main focus of toxic bot comments detected in January 2025. Different colors represent clusters of words frequently used together, revealing the bot’s primary narratives. In this case, keywords like canadian, karmila harris and carney frequently appeared together. Also, muslims, trash and biden form another toxic bot narrative.