Long-term monitoring of the evolution of the artificial night sky brightness is a key tool for developing science-informed public policies and assessing the efficacy of light pollution mitigation measures. Detecting the underlying artificial brightness trend is a challenging task, since the typical night sky brightness signal shows a large variability with characteristic time scales ranging from seconds to years. In order to effectively isolate the weak signature of the effect of interest, determining the potential long term drifts of the radiance sensing systems is crucial. If these drifts can be adequately characterized, the raw measurements could be easily corrected for them and transformed to a consistent scale. In this short note we report on the progressive darkening of the signal recorded by SQM detectors belonging to several monitoring networks, permanently installed outdoors for periods ranging from several months to several years. The sensitivity drifts were estimated by means of parallel measurements made at the beginning and at the end of the evaluation periods using reference detectors of the same kind that were little or no exposed to weathering in the intervening time. Our preliminary results suggest that SQM detectors installed outdoors steadily increase their readings at an average rate of +0.034 magSQM/arcsec2 per MWh/m2 of exposure to solar horizontal global irradiation, that for our locations translates into approximately +0.05 to +0.06 magSQM/arcsec^2 per year
