October 2, 2026
OpenAI has added a virtual try-on feature to ChatGPT, letting users see how clothing looks on them using the platform's latest image generation model. Items that catch a user's eye can be saved for later reference. The feature brings AI-assisted shopping one step closer to a practical everyday tool.
Midjourney's alpha site picked up several quality-of-life updates last week, including style previews in the sidebar, larger image display in the create view, and the ability to set default parameters at the folder level. The changes are focused on reducing guesswork and giving users more control before and during generation.
Black Forest Labs has released Flux 3 Image, the image generation side of its new Flux 3 model family. The model introduces multi-step editing that confines changes to targeted regions, leaving the rest of a composition intact. Open weights are expected to follow within the next few weeks.
A court of appeals has ordered a new sentencing hearing for a man convicted of manslaughter, after finding that the original judge was improperly influenced by an AI-generated video of the victim. The case marks a notable moment in the legal system's ongoing reckoning with generative AI as evidence. It raises serious questions about what role synthetic media should be permitted to play in courtroom proceedings.
