•A framework is proposed for the wide area automated detection of active geohazards.•The first inventory of active geohazards (AGs) in the Hexi Corridor is established.•Elevation, temperature and precipitation are the primary conditioning factors of AGs.•Faults have more control on very slow-moving landslides than on slow-moving ones.•Dual conditioning factor interaction contributes to a bivariate enhancement of AGs. With the escalation of global climate change and human activity, geohazards become increasingly frequent which cause severe casualties and property losses to local communities. To alleviate this situation and provide scientific guidance for risk reduction, it is imperative to address some of the basic questions related to geohazards, including: i) how to detect active geohazards (AGs) rapidly and automatically over a wide area; ii) how to determine the region with a high level of hazard activity; iii) what are the primary conditioning factors (CFs) of AGs; and iv) do factors operate independently or are they interconnected. To tackle these issues, we propose a universally applicable framework for wide area automated detection of AGs. The framework is based on multi-source Earth observations which capture surface deformation ranging from millimeters to meters. Our study has focused on the Hexi Corridor (HXC) in Gansu Province, China, covering an area of 210,000 km2 with a length of 1100 km. First, we construct an AGs database for the HXC with high automatic and rapid update capabilities, including a total of 4492 AGs (3652 active landslides and 840 land subsidence areas). Second, using the Geographic Detectors method, we determine the primary CFs including elevation, land surface temperature, and precipitation. We find that faults exert greater control over very slow-moving landslides, but are less effective over slow-moving landslides. Third, we analyzed the interactive effects of dual CFs on geohazard actives. Any interaction effect of dual CFs contributes to the bivariate enhancement of geohazard activity. This study significantly enhances the capabilities of the wide area automated detection of AGs, and provides a crucial dataset for hazard prediction and mitigation along the HXC.