Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.
By July 2026, a narrative review of 40 studies from 2020-2025 found deep learning, led by U-Net variants with residual and attention mechanisms and standardized pipelines like nnU-Net, increasingly achieved high Dice scores on MRI DWI/ADC, with many reports above 0.80 and recent transformer and ensemble multisite models approaching 0.90, while CT performance was lower and more variable.
In a 2022-2024 study of 472 emergency department patients with suspected pulmonary thromboembolism across three Mashhad University of Medical Sciences centers, researchers developed explainable machine-learning models intended to operate after Wells/Geneva triage. Using CTPA as reference, Extra Trees achieved accuracy 0.82, sensitivity 0.69, specificity 0.86 and AUC 0.83 for PTE prediction, and AUC 0.77 for central and 0.67 for peripheral emboli for anatomical classification.
A July 2026 peer-reviewed study compared ChatGPT GPT-5.1, Gemini 2.5 Flash, and Claude Sonnet 4.5 against expert periodontologists on 20 periodontal patient questions. Nine blinded periodontologists rated anonymized answers for scientific accuracy, completeness, conciseness & focus, empathy, and clarity.
In 282 patients with Lenke 1/2 adolescent idiopathic scoliosis treated with selective posterior thoracic fusion, investigators built an interpretable machine learning pipeline to stratify risk of postoperative coronal imbalance. After reducing 24 candidates to key features, a LightGBM model achieved the highest discrimination with AUC 0.885 in training and 0.824 in internal validation, with SHAP highlighting LIV-LSTV relationship, Lumbar Modifier, and Risser grade.
In October, a sixth-grade student at a middle school in Brooklyn received an assignment to create a science experiment and then ask Google Gemini for feedback. His mother, Kelly Clancy, objected to the teacher and later founded Parents for AI Caution in Educational Spaces, which is pushing for a two-year moratorium on AI in New York City public schools.
UK Finance reported almost 15,000 investment scams in 2025, with about £221.5m lost after people were persuaded to move money to fake investments or fictitious funds, a 40% rise on the prior year. The report said criminals use artificial intelligence to dupe people and that advances in AI make large-scale scams easier, with schemes involving gold, cryptocurrencies and wine.
At Wimbledon in July 2025, organizers replaced 300 human line judges with an artificial intelligence electronic line-judging system. Shortly after deployment, the new system failed to detect that player Sonay Kartal had hit a ball long during a match.
Cifas, the UK's fraud prevention organisation, reported 444,000 fraud cases from its members last year, a 6% rise on 2024 and a record for its national fraud database. The body said criminals are increasingly exploiting AI technology to take over mobile, banking and online shopping accounts, using stolen data to make unauthorised transactions and enabling deception on industrialised levels.
ace