All traces

Fake videos about GTA 6 generated by artificial intelligence were circulating on online platforms, prompting copyright enforcement.

Ability to design complete viral genomes raises new biosecurity risks around biological design.

Analysis of 125 compensated orthodontic cases found the majority of patient harms were avoidable and driven by multiple interacting human factors such as monitoring and clinical knowledge and organizational factors such as protocols and workplace culture.

AI-designed functional viral genomes raise biosafety concerns flagged as urgent.

AI image generators like Midjourney reproduced dataset biases, including generic ethnic features, gendered nudity assumptions, and advertising-driven distortions in recreations of famous photographs.

AI analysis of two museum paintings attributed to Jan van Eyck found no detectable Van Eyck brushstrokes, casting doubt on their attribution and creating reputational risk for the holding museums.

AI-powered search engines and models are delivering answers directly to users through zero-click searches, reducing traffic to news sites and disrupting publisher revenue models in Australia.

Participating writers worried that AI assistance could weaken creative control, interrupt their established process or reproduce a fixed version of their style rather than support continued experimentation and growth.

Fully automated AI generation of news risks bias, manipulation, lack of accountability, inaccuracy, and erosion of accuracy, trust, and copyright principles.

Personal-data-driven algorithmic targeting and other manipulative design features like infinite scrolling directly cause primary harms of manipulation, including hijacking attention, overriding user autonomy, and engineering addiction.

The same system showed substantially weaker longer-range referential and narrative continuity, lacking the sustained plotting and thematic organisation observed in Orwell's novel.

Tacit rules are negatively associated with AI autonomy feasibility, constraining full replacement to 2.7% of tasks and leaving 11.0% AI immune in cultural and creative occupations.

ChatGPT-5.0 did not achieve the highest overall classification accuracy and had lower specificity than XGBoost, indicating it missed the top performance for correctly identifying non-extraction cases.

Machine learning's black-box nature has hindered clinical adoption of mortality prediction models for non-dialysis CKD despite their clinical importance.

Patients classified as Severe profile with early onset, 73.3% co-occurring psychiatric conditions and intense craving had poorer 3-month outcomes with only 36.7% abstinence and steep return to use.