In this study, we constructed a set of scenarios for land use and land cover (LULC) change in the Colombian Amazon by 2040 using coherent narratives within the framework of the current, post-conflict stage. Three scenarios were designed: trend, extractivist, and sustainable development. The transitions between land covers in the years 2002 and 2016 and their relationship with 29 driving factors were analyzed using Artificial Neural Networks. The Markov model was calculated for the transitions, and the change allocation model was parameterized to spatially simulate the scenarios. The results showed that the LULC model calibration and validation were satisfactory (0.91). The sustainable development scenario considered strong policies for the conservation of forests and implementation of sustainable production projects to encourage the conversion of agricultural landscapes to multifunctional landscapes in recovery. As a result, areas with conserved forests and secondary vegetation in recovery could add ~ 42 million hectares by 2040. The other scenarios show that the Colombian Amazon will lose ~ 2 million hectares of forests in the trend scenario and ~ 4.3 million hectares in the extractivist scenario with respect to the reference year (2016). In the trend scenario, pastures and crops could increase by 48%, and, in the extractivist scenario, they would increase by 117%, going from ~ 3.9 to ~ 8.6 million hectares. We hope that the scientific contribution of this study will be relevant for informed discussion in decision-making and provide a framework for building a peaceful territory.