October 7, 2026
OpenAI has introduced a visual advertising format that displays ads alongside image generation results, starting with users in the United States. The move marks a notable step toward ad-based revenue for the company, which has historically relied on subscriptions and API access. It is currently a limited rollout, with no confirmed timeline for expansion to other regions.
Reka has released Rho-1, a 19-billion-parameter model that handles text, images, video understanding, video generation, and robot action outputs within a single unified network. The architecture uses a shared key-value cache across all modalities, and a distilled variant can produce a 5.3-second video clip in roughly one second. The release is currently a research preview, with no public weights available.
An AI-generated video of a murder victim was used as a victim impact statement in an Arizona court, with the deceased man's sister creating an avatar to speak in his place. While the judge was visibly moved by the presentation, an appeals court later found that the video carried "undue emotional weight" - raising serious questions about the role of generative AI in legal proceedings.
Google has released Nano Banana 2.1, a new image generation model that uses the Gemini 3.6 Flash backend and undercuts the cost of its predecessor while matching or exceeding it on several benchmarks. The improvement comes with a caveat, though - the previous model also tested well, and the Pro tier has historically delivered stronger results in real-world use.
Researchers have developed an AI model capable of reconstructing images from a person's brain activity with notable speed and accuracy. The system interprets neural signals and generates visual outputs that closely match what a subject is actually viewing. The results mark a meaningful step forward in the field of brain-computer interfaces and neural decoding.
Shengshu has released Vidu Q4, the latest version of its AI video generation model, adding native audio generation alongside improvements to motion quality and prompt adherence. The update positions Vidu Q4 as a more complete tool for creative video work, handling both visual and audio output within a single model. These changes reflect a broader trend in the field toward unified multimodal video generation.
Maket has released version 2.0 of its AI-powered architectural design tool, adding the ability to generate floor plans and 3D home renders directly from prompts. The update marks a significant expansion of the platform's capabilities beyond its earlier interior design focus. Architects, developers, and homeowners now have a more complete design workflow within a single tool.
Google has launched a public-facing website for SynthID, its AI watermarking technology, allowing anyone to upload and check whether an image, video, or audio clip was generated by AI. The tool extends SynthID beyond its previous developer and enterprise integrations, putting detection capabilities directly in the hands of everyday users. This marks a notable step toward accessible, practical tools for identifying synthetic media in the wild.
