In a global computing environment in order for entities to collaborate, they should be able to make autonomous access control decisions with partial information about their potential collaborators. The SECURE project addresses this requirement by using trust as the mechanism for managing risks and uncertainty. This paper describes how trust lifecycle management, a procedure of collecting and processing evidence, is used by the SECURE collaboration model. Particular emphasis is placed on the processing of the evidence and the notion of attraction. Attraction considers the effects of evidence about the behaviour of a particular principal on its current trust value both in terms of trustworthiness and certainty and is one of the distinctive characteristics of the SECURE collaboration making it more appropriate for a global computing setting.