In everyday interactions we find our attention follows the eye gaze of faces around us. As this cueing is so powerful and difficult to inhibit, gaze can therefore be used to facilitate or disrupt visual processing of the environment, and when we experience this facilitation or disruption ourselves we infer information about the cueing face, specifically manifesting in judgements of trustworthiness. However, to date no studies have investigated how resilient these impressions are to interference, or how long they last. To explore these questions we used a gaze-cueing paradigm where 16 faces demonstrated either valid (always cueing the correct location) or invalid (always cueing the incorrect location) cueing behaviours. Previous experiments using this paradigm show that valid faces are subsequently rated as more trustworthy than invalid faces. In answer to the first question we show, when introducing a short interference task between gaze-cueing and trustworthiness ratings, this effect is extinguished. However, if the face identities are initially encoded deeply this enables more stable learning of the association between identity and gaze patterns. Addressing the second question we find the effect is weaker but that the overall pattern of results.