Usual structural hypothesis related to superionic glasses containing metal halides MY = LiI, NaBr, AgI, etc., implies the absence of chemical interactions between the host glass and metal halide. An expansion of the host network caused by large anions (I, Br) and related increase of the available free volume for fast cations, forming preferential conduction pathways, was considered to be the key parameter for glassy superionics. Our preliminary results for a (AgI)0.2(As2S3)0.8 glass show that this AgI-poor sample exhibits a mixed silver tetrahedral environment (2S+2I). Consequently, a strong evidence of chemical interaction between AgI and As2S3 host glass is observed. We are asking for 2 days on GEM or SANDALS to elucidate the short-range order and connectivity of the AgI-related subnetwork and follow the expected evolution from isolated to extended Ag-centred structures.