The next speaker in this session at the ECREA 2026 conference in Brno is Konstiantyn Yanchenko, whose focus is on social actors on Telegram. In Europe, Telegram is used as a source of news specially in southern countries (Spain, Italy, Serbia, Turkey), and features a degree of institutionalisation: there are news outlets, government institutions, politicians, as well as influencers and individual, sometimes anonymous users; in democracies as well as autocracies, it also provides a platform for alternative and dissident voices.
What identity cues do Telegram actors provide, then? What external information is available about them? Do these match the actual identity of actors running these channels? This project explored this using a snowball crawl starting with 40 seed channels, whose messages were captured through to the start of 2025, and examined for forwards from other channels; these channels were then also captured and analysed in a similar way. These channels were filtered for eight key languages, and for public availability, and this produced a total set of some 60,000 channels, with more than half of them posting in Russian.
Automated content analysis then examined these channels for the presence or absence of actor functionalisation (absent / individual / collective), and for their actor group (politics, journalism, civil society, etc.). Most channels (up to 88% in Portuguese) provided no explicit information whatsoever; Ukrainian and Russian channels were most explicit. Those languages also had the greatest percentage of collective channels.
Media and politics channels accounted for the greatest share; military channels were especially prominent in Ukrainian, and finance and culture channels especially in Russian; civil society in Italian. Media and finance channels tended to have the largest number of followers, politics and civil society channels the fewest (for politics, with the exception of Ukraine). This may also be because political actors do not necessarily self-describe as politicians, however.
