Context. We study upflows at the edges of active regions, called AR outflows, using multi-instrument observations. Aims. This study intends to provide the first direct observational evidence of whether chromospheric jets play an important role in furnishing mass that could sustain coronal upflows. The evolution of the photospheric magnetic field, associated with the footpoints of the upflow region and the plasma properties of active region upflows is investigated with the aim of providing information for benchmarking data-driven modelling of this solar feature. Methods. We spatially and temporally combine multi-instrument observations obtained with the Extreme-ultraviolet Imaging Spectrometer on board the Hinode, the Atmospheric Imaging Assembly and the Helioseismic Magnetic Imager instruments on board the Solar Dynamics Observatory and the Interferometric BI-dimensional Spectro-polarimeter installed at the National Solar Observatory, Sac Peak, to study the plasma parameters of the upflows and the impact of the chromosphere on active region upflows. Results. Our analysis shows that the studied active region upflow presents similarly to those studied previously, i.e. it displays blueshifted emission of 5-20 km s −1 in Fe xii and Fe xiii and its average electron density is 1.8 × 10 9 cm −3 at 1 MK. The time variation of the density is obtained showing no significant change (in a 3σ error). The plasma density along a single loop is calculated revealing a drop of 50% over a distance of ∼20 000 km along the loop. We find a second velocity component in the blue wing of the Fe xii and Fe xiii lines at 105 km s −1 reported only once before. For the first time we study the time evolution of this component at high cadence and find that it is persistent during the whole observing period of 3.5 h with variations of only ±15 km s −1 . We also, for the first time, study the evolution of the photospheric magnetic field at high cadence and find that magnetic flux diffusion is responsible for the formation of the upflow region. High cadence Hα observations are used to study the chromosphere at the footpoints of the upflow region. We find no significant jet-like (spicule/rapid blue excursion) activity to account for several hours/days of plasma upflow. The jet-like activity in this region is not continuous and blueward asymmetries are a bare minimum. Using an image enhancement technique for imaging and spectral data, we show that the coronal structures seen in the AIA 193 Å channel are comparable to the EIS Fe xii images, while images in the AIA 171 Å channel reveal additional loops that are a result of contribution from cooler emission to this channel. Conclusions. Our results suggest that at chromospheric heights there are no signatures that support the possible contribution of spicules to active region upflows. We suggest that magnetic flux diffusion is responsible for the formation of the coronal upflows. The existence of two velocity components possibly indicates the presence of two different flows, which are produce...