Empirical studies of the principal-agent relationship find that extrinsic incentives work in many instances, linking rewards to performance increases effort, but that they can also backfire, reducing effort. Intrinsic motivation, the internal drive to work to master a skill or to improve one's self image, is thought to be the key to whether incentives work or not. If the incentives crowd-out intrinsic motivation, and the effect is large enough, the net motivational effect on effort will be negative. We posit that an aversion to being exploited, i.e. being used instrumentally for the benefit of another, is one facet of intrinsic motivation, triggered by the combination of high-powered incentives and egoistic principal intent, that can cause incentives to fail. Using an experiment that provides the material circumstances necessary for exploitation to occur, we find that agent compliance is significantly lower for exploitative principals who use high-powered incentives and have a financial interest to do so, compared to neutral principals who use the same contracts but do not benefit from them. To corroborate our interpretation of the results we show that a surveyed "exploitation aversion" scale moderates this effect. Exploitation averse participants are less likely to comply with the incentives than exploitation tolerant participants when the principal signals an exploitative intent, but they are no less likely to comply with the same incentives when the principal is neutral. Our results have implications for the design and implementation of incentive structures within organizations.