Method for automatically protecting an object, a person, or an item of visual information from a risk of unwanted viewing by a third party, an unauthorized person and/or an image-recording device. The method includes the steps of: automatically detecting, by analyzing images from at least one Light Detection and Ranging (LIDAR) or from a camera arranged so as to cover a predefined surveillance area, the presence of a third party or of an unauthorized person and/or of the device in this area or nearby, the predefined surveillance area corresponding at least to an area from which the person or the device is capable of viewing the object, the person or the information, and in the event of positive detection, triggering a predefined action.