In March 2023, Grant was preparing to teach his general entomology class when a student sauntered in and said that he had a photograph of an insect he couldn’t identify. Grant considers himself a fairly good insect taxonomist and can usually put a family name on most of the major groups in the USA. When he glanced at the image, though, something didn’t ring true: he’d never seen any insect that looked quite like it. A beetle, for sure, but what kind? Grant told the student that he must not have taken the photograph in their part of the world (northwest Florida). The student told Grant that indeed he hadn’t. Instead, he’d asked an AI image generator to create it and, moments later, he had a “deepfake” insect photo. While this particular photo is now lost to the digital ether, Figure 1 offers some examples made with getimg.ai. Deepfakes are AI-produced media that appear authentic. Westerlund (2019) and Seow et al. (2022) reviewed the emergence of deepfake technology, anticipating that it will soon be nearly impossible for humans to distinguish deepfakes from true media.
No takes yet. Share an insight, caveat, or question.
Jong et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: