Rese­arch and socie­ty seem to agree that Gene­ra­ti­ve AI is about to pro­found­ly influence our modes of com­mu­ni­ca­ti­on, yet its pre­cise impact remains ambi­guous. Is Tris­tan Har­ris and Ira Raskin’s (Cen­ter of Huma­ne Tech­no­lo­gy) asser­ti­on true that what nuclear wea­pons were to the phy­si­cal world, gene­ra­ti­ve AI is to the vir­tu­al and sym­bo­lic world? Or should we rather aim to rethink the ine­vi­ta­bi­li­ty of AI? Does gene­ra­ti­ve AI genui­ne­ly chall­enge tra­di­tio­nal modes, chan­nels, and media of com­mu­ni­ca­ti­on, or is it sim­ply a con­tri­van­ce to bols­ter the pro­files and pro­fits of tech con­glo­me­ra­tes? Bet­ween the poles of this spec­trum, the­re remains a wide ran­ge of inter­pre­ta­ti­ons deman­ding nuan­ced analysis.

Gras­ping the full scope of gene­ra­ti­ve AI’s impact is thus cri­ti­cal. As gene­ra­ti­ve AI tools beco­me capa­ble of pro­du­cing or co-crea­ting it, the very noti­on of ‘con­tent’ is evol­ving – lea­ding to deba­tes over aut­hor­ship, par­ro­tism and copy­right. Impli­ca­ti­ons of a rhe­to­ri­cal­ly acting AI that con­vin­cin­g­ly mimics human out­put chall­enge us deep­ly, spar­king dis­cus­sions about ‘arti­fi­ci­al influence’. With such a per­sua­si­ve AI’s poten­ti­al to dis­rupt power dyna­mics, espe­ci­al­ly in poli­ti­cal are­nas, its role in the 2024 glo­bal elections—particularly the US elec­tion in Novem­ber—will ser­ve as a cri­ti­cal lit­mus test.

Rhe­to­ri­cal ana­ly­sis has long aimed to under­stand and ethi­cal­ly navi­ga­te the bia­ses and moti­ves of com­mu­ni­ca­ti­on. Gene­ra­ti­ve AI intro­du­ces new pro­blems to this chall­enge, yet the core issue per­sists: dis­cer­ning intent, inte­rests and moti­va­tions. As AI-gene­ra­ted con­tent beco­mes more sophisti­ca­ted, our rhe­to­ri­cal veri­fi­ca­ti­on methods must also evol­ve, par­ti­cu­lar­ly to address pos­si­ble mali­cious human-AI collaborations.