How are AI-generated images detected in photography competitions?

Numbers match the citations in the article. A working link isn't proof that a page supports a claim; check the quoted passage and date.

1. 1

   Biomedical Research Images Manipulated by Generative AI to Alter Scientific Outcomes: Diagnostic Study of Human and Automated Detection.

   Journal of medical Internet research (Luo et al.)Published Oct 2, 2026Checked Oct 11, 2026

   “UnlabelledIn a diagnostic study of 104 western blot and subcutaneous xenograft tumor images, a high-fidelity generative model produced forgeries that could alter the conclusions of a study; 24 PhD-level expert reviewers could not reliably distinguish the forgeries from authentic figures (mean accuracy 50.5%, SD 6.9%), while the best-performing commercial AI detector achieved only moderate discrimination (area under the curve 0.790, 95% CI 0.695-0.885), revealing critical vulnerabilities in current research-integrity safeguards.”

2. 2

   Evaluating the use of a novel expert system for interpretation of media authentication results.

   Journal of forensic sciences (Epstein et al.)Published Jun 11, 2026Checked Oct 11, 2026

   “These tests were organized into distinct authentication pathways to produce concise, natural language conclusions comparable to opinions rendered by human examiners. The automated nature of an expert system may allow for demonstrably accurate results at scale without extensive human resources. This preliminary study compared the accuracy of the artificial intelligence expert system's opinions with those of trained human examiners from diverse digital forensic backgrounds. Both the expert system and the examiners were provided the same dataset, consisting of original, transmitted, edited, and synthetic videos. Both groups received file structure data, attribute similarity analyses, proprietary structural data, and metadata values. Using a 14-question survey, responses were collected and evaluated for accuracy, and error rates were calculated. Across 20 respondents, the human mean score was 69% as compared to the expert system score of 91%. These results can be used to evaluate the appropriateness of deploying artificial intelligence expert systems for forensic examinations, alongside human interpretations.”

3. 3

   Penny-Wise and Pound-Foolish in AI-Generated Image Detection.

   IEEE transactions on pattern analysis and machine intelligence (Wang et al.)Published Jul 1, 2026Checked Oct 11, 2026

   “To address this trade-off issue, we propose a novel learning framework (PoundNet) for the generalization of AI-generated image detection on a pre-trained vision-language model. PoundNet incorporates a learnable prompt design and a balanced objective to preserve broad knowledge from upstream tasks (object classification) while enhancing generalization for downstream tasks (AI-generated image detection). We train PoundNet on a single standard AI image dataset, following common practice in the literature. We then evaluate its performance across 10 large-scale public AI-generated image detection datasets with 5 main evaluation metrics, forming the largest benchmark test set for assessing the generalization ability of AI-generated image detection models, to our knowledge. The comprehensive benchmark evaluation demonstrates that PoundNet successfully balances generalization with knowledge retention, achieving a remarkable relative improvement of 19% in AI-generated image detection performance compared to state-of-the-art methods, while maintaining a strong performance of 63% on object classification tasks.”

4. 4

   Interpol review of detection of AI-generated image and video deepfakes, 2022-2025.

   Forensic science international. Synergy (van et al.)Published Jun 13, 2026Checked Oct 11, 2026

   “We analyzed more than 400 studies published between 2022 and 2025, specifically addressing the generation and detection of AI-generated deepfakes as opposed to manipulation of existing media. We provide an overview of the developments in various generative architectures such as GANs, diffusion models, and autoregressive models, highlighting progress in visual fidelity, user control, content consistency, and computational efficiency. In addition, we outline the most common designs of detection methods, looking at various types of features that are used for detection. We conclude that innovations are mainly centered on addressing the challenges of generalizability to unseen generators and robustness against common perturbations and adversarial attacks. However, the inconsistent use of the evaluation datasets makes it difficult to compare the methods. Our review has identified several directions for future research, such as making methods more directly applicable to forensic practice by incorporating forensically relevant benchmark datasets, paying more attention to explainability, and embedding detection methods in an evidence evaluation framework.”

5. 5

   Prize-winning image which sparked backlash was AI-generated, Nikon rules

   BBC NewsPublished Oct 9, 2026Checked Oct 11, 2026

   “The camera-maker says it is now re-evaluating the rules and procedures of its Small World in Motion contest.”

6. 6

   CoDA: Color Distribution Probing for Efficient and Generalizable AI-Generated Image Detection.

   IEEE transactions on pattern analysis and machine intelligence (Jia et al.)Published Oct 1, 2026Checked Oct 11, 2026

   “Motivated by this broader setting, we revisit color-distribution probing as an efficient complementary cue for AI-generated image detection. We observe that, especially for photographic content, real photographs tend to exhibit smoother and more stable color patterns, whereas synthetic images often show characteristic color imbalances introduced by neural generation. Based on this observation, we propose CoDA, a compact 1.48M-parameter detector built on a Noise-Quantization Probe, together with a theoretical analysis linking probe responses to color non-uniformity. Experiments show that CoDA achieves state-of-the-art performance on standard benchmarks and the best results on the challenging cross-domain evaluation of FakeForm, while remaining highly competitive in cross-model photorealistic settings. These results suggest that persistent generative artifacts can provide a practical foundation for efficient and robust AI-generated image detection. The models and FakeForm benchmark will be made publicly available.”