Contrary to the standard lore, there is mounting observational evidence that feedback from active galactic nuclei (AGN) may also play a role at the low-mass end of the galaxy population. In this talk, I will explore this possibility employing both isolated and cosmological simulations of dwarf galaxies. Firstly, I will present the results from my high-resolution simulations with the moving mesh code AREPO. In these isolated dwarf simulations, the AGN has only a limited direct effect on star formation rates. There is, however, a significant effect on outflows, which are notably enhanced by the AGN to much higher temperatures and velocities. This indicates that AGN may play an indirect role in quenching dwarf galaxies by hindering cosmic gas inflows. I further investigate this quenching scenario using the cosmological simulation suite FABLE. While in the local Universe the majority of AGN in dwarfs are much dimmer than the stellar component, for z ≥ 2 there is a significant population that outshines their hosts. These high-redshift overmassive BHs contribute to the quenching of dwarfs, whereas at late cosmic times supernova (SN) feedback is more efficient. However, the lack of high-luminosity X-ray AGN in FABLE at low redshifts highlights an interesting possibility that SN feedback could be too strong in FABLE’s dwarfs, curtailing AGN growth and feedback. Using a series of zoom-in simulations, varying SN energetics, I examine the interplay between AGN and SN feedback at the low-mass end and I will discuss the necessary conditions for AGN in dwarfs to reach high accretion rates. Finally, I will also consider the impact of the AGN accretion model itself and present a new unified accretion disc model, combining the ADIOS flow and standard thin disc schemes.