Free Speech for Bots and Trolls? Foreign Interference and Democratic Self-Defense

(This paper could be included in Panel 2: "Epistemic integrity and the problem of misinformation") My goal in this paper is to discuss whether foreign interference, particularly the use of bots and trolls to influence the public debate in social media (Morrison 2021, Prodan & Gombar 2025), could be considered incompatible with democracy. Such interferences exploit the very principles of liberal democracy, such as free speech and pluralism (Wigell 2019), thus measures to counter them pose a dilemma for democratic self-defense. Moreover, not all foreign influences are a threat to democracy, some may even have a positive effect on the public debate. I propose to use the concept of manipulation to identify illegitimate foreign influences. Theorists of democracy usually frame manipulation as a problem, even though there is no agreement on its definition (Sustein 2015, Withfield 2022). In general terms, in a democracy citizens are expected to freely express their true interests, which are then aggregated according to certain procedures. If citizens are manipulated, they do not have access to their true interests, so the democratic process is compromised. I define non-manipulative speech as requiring three conditions: transparency (the identity and interests of a source should not be falsified), non-malicious intentions (the speech should not be willfully designed to damage the interests of the target) and sincerity (which excludes intentional falsity or ambiguity). By this definition, the use of bots and trolls to influence the public debate may count as manipulation. First, the condition of transparency is violated when bots and trolls controlled by a foreign actor are disguised as real citizens. Second, malicious intentions can be assumed when bots and trolls are controlled by a hostile foreign actor. Third, bots and trolls do not engage with sincerity, since they have no interest in correcting false statements or solving ambiguities when these serve their narrative. These criteria are enough to identify and exclude manipulative foreign interferences. However, the issue is complicated by the fact that real citizens who have already been manipulated will share their views. These manipulated citizens play the same role as bots and trolls, but their action may not count as manipulation: they respect transparency, it cannot be proved that they have malicious intentions, and it cannot be proved that their false or ambiguous claims are intentionally false or ambiguous. This suggests that democratic self-defense must go beyond protection from foreign interferences, and that, regardless of manipulation, we also have to find ways to engage with citizens who have ideas that may be incompatible with democratic principles.