OpenAI and Google just slashed free usage limits for Sora and Nano Banana Pro after surging holiday demand overwhelmed their compute. Sora free users are now capped at six video generations a day, while Google reduced image generations to two daily and restricted Gemini 3 Pro access. Basically, running frontier video and image models is so expensive that even the tech giants are feeling the squeeze.
LTX Studio just rolled out Retake, letting creators direct and tweak AI video shots even after they have finished rendering. Users can now rephrase dialogue, reshape character emotions, and reframe cameras without redoing the entire clip. Why it matters: it makes fine-tuning AI videos far less painful and costly.
OpenAI just got hit with a temporary restraining order blocking it from using the word "cameo" inside its Sora app. Celebrity video platform Cameo claimed trademark infringement over Sora’s likeness-swapping feature, and a federal judge agreed to pause OpenAI's usage through December 22. OpenAI is pushing back, arguing no single company can exclusively own the everyday term.
Twelve Labs just integrated its Marengo and Pegasus multimodal models directly into Adobe's Frame.io V4 via Custom Actions. Creative teams can now right-click to run semantic searches, find B-roll with image queries, auto-generate metadata, and flag compliance issues on the timeline without leaving their review workflow.
Amazon just rolled out "Video Recaps" on Prime Video, using generative AI to stitch together key clips, dialogue, music, and an AI voiceover into short catch-up videos. It is currently in beta for US living room devices on a handful of originals like Fallout. Why it matters: You no longer need to read lengthy wiki summaries or watch 20-minute fan recaps before starting a new season.
CraftStory, an AI startup founded by OpenCV's creators, just emerged from stealth with $2 million in funding and Model 2.0, an AI system that generates coherent human videos up to five minutes long. Using a parallelized diffusion architecture, it processes full video durations simultaneously to prevent accumulated visual artifacts. Why it matters: Long-form AI video has remained elusive, and CraftStory aims to turn AI clips into viable full-length tutorials and marketing demos.
TikTok just unveiled new controls that let users adjust how much AI-generated content pops up in their For You feeds. It is also testing robust invisible watermarks to stop AI labels from being stripped during re-uploads, alongside launching a $2M educational fund for responsible AI literacy.
Google just opened up several core Gemini AI features in Google Vids to anyone with a standard Gmail account, removing the paid Workspace paywall. Creators can now tap AI-generated voiceovers, transcript-based trimming to clear filler words and silences, and built-in AI image editing for free.
ElevenLabs just launched ElevenLabs Image & Video in beta, turning its voice platform into an all-in-one multimodal creative suite. Users can now generate visuals using models like Sora, Veo, Kling, and Flux, sync them with voiceovers and sound effects, and polish everything inside its Studio editor. It is a big push toward unifying visual and audio creation in a single workflow.
A former Twitter employee just launched diVine, a Vine reboot backed by Jack Dorsey that takes a strict stance against generative AI. Built on the Nostr protocol, the platform has restored 170,000 vintage loops and strictly limits content to human-made clips. It is a deliberate bet that users are growing tired of synthetic social feeds.
NotebookLM just launched custom video overview styles, letting users dictate the look and tone of their AI video summaries simply by typing a prompt into the customization box. Why it matters: It moves the tool beyond rigid templates, giving users direct creative control over automated summaries.
Fei-Fei Li, Yann LeCun, and DeepMind are all heavily betting on "world models," but they basically mean three completely different architectures. World Labs' Marble delivers interactive 3D Gaussian splats, LeCun's JEPA builds an internal predictive latent brain, and DeepMind's Genie 3 and SIMA 2 generate real-time video simulators. Why it matters: The industry is packaging front-end assets, robot gymnasiums, and core machine cognition under a single marketing buzzword.
Epidemic Sound just launched Studio, an AI-powered tool that automatically builds synchronized soundtracks for uploaded videos using licensed music and foley effects. It draws on data from billions of daily plays to match audio to visuals, aiming to turn a multi-hour scoring grind into a quick, copyright-safe process.
Google's Flow platform just launched a new "Images" tab in the top left of projects. The dedicated workspace lets creators generate and refine visuals with Nano Banana separately from video outputs. Why it matters: Separating asset refinement from final video generation keeps projects organized when fine-tuning creative ideas.
Higgsfield AI just launched Recast, a creative feature built for advanced face and character swapping in videos and images. The tool aims to move past simple gimmick swaps by tuning face shapes, skin tones, and perspective angles. That said, hitting production-ready quality still takes manual masking and multi-pass work.
Meta just brought its revamped Meta AI app to Europe, rolling out Vibes—a TikTok-like feed dedicated to AI-generated short videos. Users can generate clips from prompts, remix other creators' posts, add music, and push videos directly to Reels or Stories, while also managing Ray-Ban Meta smart glasses in the same hub.
Netflix just told investors it is leaning heavily into generative AI, using the technology for special effects, de-aging actors, and pre-production design. CEO Ted Sarandos made it clear that AI is strictly an efficiency tool for artists rather than a replacement for human storytelling, insisting the studio is not worried about losing the human touch.
OpenAI just issued a joint apology after Sora 2 generated unauthorized deepfakes of Breaking Bad star Bryan Cranston, including one posing with Michael Jackson. The company promised to strengthen guardrails around voice and likeness opt-ins, while SAG-AFTRA is using the incident to push for federal anti-cloning legislation.
Krea AI just open-sourced Krea Realtime, a 14B autoregressive model that is 10x larger than existing open-source counterparts. Running on a single B200 GPU, it generates long-form video at 11 fps. Weights and a technical report are available now.
OpenAI just blocked users from generating AI videos of Martin Luther King Jr. in Sora following requests from Dr. King's estate. The move comes after users generated offensive and disrespectful depictions of the civil rights icon on the newly launched app. OpenAI confirmed that authorized representatives of public figures can now opt out of Sora cameos.
