ABSTRACT X-ray luminosity functions (XLFs) of active galactic nuclei (AGNs) trace the growth and evolution of supermassive black hole populations across cosmic time. However, current XLF models are poorly constrained at redshifts of z > 6, with a lack of spectroscopic constraints at these high redshifts. In this work, we place limits on the bright-end of the XLF at z = 5.7–6.4 using high-redshift AGN identified within the Extragalactic Serendipitous Swift Survey (ExSeSS) catalogue. Within ExSeSS, we find one serendipitously X-ray detected z > 6 AGN, ATLAS J025.6821-33.4627, with an X-ray luminosity of $L_\mathrm{X}=8.47^{+3.40}_{-3.13}\times 10^{44}\mathrm{erg.s^{-1}}$ and z = 6.31 ± 0.03, making it the highest redshift, spectroscopically confirmed, serendipitously X-ray detected quasar known to date. We also calculate an upper limit on the space density at higher luminosities where no additional sources are found, enabling us to place constraints on the shape of the XLF. Our results are consistent with the rapid decline in the space densities of high-luminosity AGN towards high redshift as predicted by extrapolations of existing parametric models of the XLF. We also find that our X-ray based measurements are consistent with estimates of the bolometric quasar luminosity function based on UV measurements at z ≳ 6, although they require a large X-ray to bolometric correction factor at these high luminosities.