Anyone can fake a scientific image with AI, tricking even academic journals – and undermining trust in science
Publish Date: 2026-06-22 08:36:00
Source Domain: theconversation.com
- The Artemis II mission photograph of Earth and the Moon has been celebrated like the Apollo 8 “Earthrise” image but raises concerns about distinguishing real scientific images from AI-generated ones.
- AI-generated images’ proliferation in scientific communication contributes to a crisis of trust in science, as traditional methods to establish visual credibility are becoming less reliable.
- AI tools are reshaping how scientific visuals are created, shared, and publicized, often blurring the line between illustration, enhancement, and fabrication, leading to retractions and concerns in certain scientific fields.
- Academic publishers are starting to adopt AI-detection tools, but these systems often lag behind newer AI creation models, particularly with sophisticated AI visuals that can subtly distort scientific details.
- Historically, scientific images carried authority because of the specialized expertise, expensive equipment, and institutional resources needed to create them. Generative AI is undermining the reliability of these traditional cues.
- AI imagery can lead to amplified motivated reasoning; genuine scientific images contradicting personal beliefs are often dismissed as AI-generated, while fabricated images supporting such beliefs are accepted.
- Authentic scientific images still hold power due to their documented connection to reality, featuring astronauts, physical equipment, and verified missions.
- Transparent, well-documented, and clearly communicated use of AI-generated scientific images is crucial to maintaining public trust in science. Without clear guidelines, science risks losing its powerful visual communication tool.
