Stellar activity is one of the main obstacles to high-precision exoplanet observations and has motivated extensive studies in detection and characterization problems. Most efforts focused on unocculted starspots in optical transit spectrophotometry, while the impact of starspot crossings is assumed to be negligible in the near-infrared. Here, we present textit{HST}/WFC3 transit observations of the active star WASP-52, hosting an inflated hot Jupiter, which present a possible starspot occultation signal. By using this data set as a benchmark, we investigated whether the masking of the transit profile distortion or modeling it with both a starspot model and a Gaussian process affects the shape of the transmission spectrum. Different methods produced spectra with the same shape and a robust detection of water vapor, and with $lesssim 1 sigma$ different reference radii for the planet. The solutions of all methods are in agreement and reached a similar level of precision. Our WFC3 light curve of WASP-52b hints that starspot crossings might become more problematic with textit{JWST}s higher sensitivity and complete coverage of the transit profile.