Despite two decades of study, the full scope of RNAi in mammalian cells has remained obscure. Here we combine: 1) Knockout of argonaute (AGO) variants; 2) RNA sequencing analysis of gene expression changes; and 3) Crosslinking Immunoprecipitation Sequencing (CLIP-seq) using anti-AGO2 antibody to identify potential microRNA (miRNA) binding sites. We find that knocking out AGO1, AGO2, and AGO3 are necessary to achieve full impact on gene expression. CLIP-seq reveals several hundred significant AGO2 associations within the 3’-untranslated regions of cytoplasmic transcripts. The standard mechanism of miRNA action would suggest that these associations repress gene expression. Contrary to this expectation, clusters are poorly correlated with gene repression in wild-type versus knockout cells. Many clusters are associated with increased gene expression in wild-type versus knock out cells, including the strongest cluster within the MYC 3’-UTR. Our results suggest that assumptions about miRNA action should be re-examined.