Geoengineering, i.e. the use of artificial techniques aiming at cooling the planet, is increasingly considered as a realistic alternative to emission mitigation. Several methods are promising for their capacity to quickly halt global warming at a moderate cost. Such cheap technologies might be very beneficial to countries profoundly affected by global warming. In this paper, I propose a dynamic model in which geoengineering is introduced as an alternative to mitigation. Contrary to abatement, geoengineering is fast and cheap, but requires a large initial investment in research and development. Within this framework, I confirm the fear which is common among geoengineering opponents : abatement is reduced if geoengineering is expected to be available in the future. The long-run implications of the model are also alarming as geoengineering will not be undertaken progressively. The sudden implementation of geoengineering, together with the sharp jump in temperature induced, may disturb climate equilibrium and fragile ecosystems. Furthermore, the availability of geoengineering will exacerbate intergenerational issues : while current generations will anticipate the use of geoengineering by increasing their emissions, future generations will have to reduce their emissions, to bear the cost of sustaining geoengineering for centuries and to suffer from its negative side-effects.