This study investigates the efficacy of collective sensing as a mechanism for unveiling disinformation in group interaction. Small group interactions were simulated to experiment on the effects of a group reaction to incentivized deceptive behavior when initiated by social influencers. We use multilevel modeling to examine the individual communication data nested within group interactions. The study advances the use of computational efficacy to support the supposition of collective sensing—by analyzing individual social actors’ communicative language and interaction within group contexts. Language-action cues as stigmergic signals were systemically extracted, compared and analyzed within groups as well as between groups. The results demonstrate that patterns of group communication become more concentrated and expressive after a social influencer becomes deceptive, even when the act of deception itself is not obvious to any individual. That is, individuals in the group characterize deceptive situations differently, but communication patterns depict the group’s ability to collectively sense deception from circulating disinformation. The study confirms our postulation of using collective sensing to detect deceptive influences in a group.