Synthetic Model Photo Sourcing

Generated Photos produces images of people who do not exist, built from a model rather than photographed, so there is no release form tied to a real person. A Beam agent reads an incoming creative request, such as a campaign brief calling for a person of a given description, and fetches a corresponding image from Generated Photos according to filters a team has approved. The image is added to the asset folder tied to that request with a note on its source. A person still reviews the final selection before it appears in a public facing piece, since fit and tone stay a judgment call.

Attribute Filtered Image Search

Generated Photos images carry attributes such as approximate age range, expression, and framing that a request can filter on. A Beam agent reads the attributes listed in a creative brief, applies the filter set a team has approved, such as headshot framing within an age range, and returns a short list of corresponding images to the asset folder rather than a single automatic pick. Nothing gets inserted into a live page or deck without a person looking at it first. A person makes the final choice from the shortlist and confirms it fits the brand and the context the image will appear in.

Source Attribution Notes

Because Generated Photos images are not tied to a real person, a team can keep them separate from licensed stock so nobody mistakes usage rights later. A Beam agent notes the source and generation reference for every image it adds to a request, and writes that detail into the asset record alongside the file. This gives a reviewer a quick way to confirm an image needs no release before it goes live. Anything unclear, such as an image that resembles a recognizable public figure too closely, gets flagged to a person instead of being filed as release free automatically.