Informacja jako broń: Od PSYOPS do algorytmów rekomendacyjnych

The article examines the thesis that recommender algorithms can perform the function of
information weapon systems. This does not mean that every system of content personalization is a weapon in a legal, military or intentional sense. It means that the recommender
infrastructure of digital platforms has features that may be exploited in influence operations: selection, prioritization, personalization, automation, amplification and feedback
loops. The article combines research on information warfare, PSYOPS, computational
propaganda, recommender systems, generative AI and the European Union debate on
the DSA, the AI Act and FIMI. The main argument is that the problem is not the mere
existence of an algorithm, but the functional connection between intentional influence and
infrastructure that automatically amplifies selected cognitive stimuli. Democratic societies
should therefore be defended not through centralized control of truth, but through limiting
manipulative infrastructure, strengthening cognitive resilience, transparency, auditability
and platform accountability.