The Academy just updated its rules to officially disqualify AI-generated performances and screenplays from Oscar contention. Acting awards now require demonstrably human performances with consent, while scripts must be human-authored, with the Academy reserving the right to audit AI usage. Hollywood is making sure digital clones won't be taking home statuettes anytime soon.
Researchers just introduced UniVid-X, a multimodal video framework that can decode and generate intrinsic properties like surface normals, lighting, albedo, and alpha transparency layers. Instead of treating video generation as flat RGB frames, it allows creators to relight scenes with custom lighting or swap characters and backgrounds with fine-grained precision.
Researchers just released Mocap Anything v2, an open-source AI framework that turns regular monocular video into clean, animatable 3D skeletons for almost any creature. Unlike older two-stage tools prone to unnatural joint twists, it jointly optimizes positions and rotations in an end-to-end model, slashing rotation errors while running 20x faster.
Nvidia just previewed D-Rex, a new framework that decouples full-body avatar animation from lighting. Instead of complex 3D reflectance math, it animates a base avatar under simple white light and lets a video diffusion model handle HDR relighting, keeping facial expressions sharp in any virtual scene.
Microsoft just introduced World-R1, a framework that enforces 3D geometric consistency in text-to-video generation using reinforcement learning. Built on Wan2.1, it injects spatial awareness without altering the base architecture, bridging the gap between video synthesis and scalable world simulation.
OmniShotCut just made video editing breakdown effortless by automatically pinpointing cut points and complex transitions like dissolves, fades, and zooms. Weighing just 164MB, the lightweight model runs easily on local hardware and is also available as a free Hugging Face demo.
Fudan researchers just unveiled Hallo-Live, a streaming framework that synthesizes synchronized talking-head video and speech directly from text. Powered by dual-stream diffusion and preference-guided distillation, it hits 20.38 FPS with a 0.94-second latency on two NVIDIA H200 GPUs. It clears a major bottleneck for interactive, real-time digital humans.
Grok Imagine just upgraded its image-to-video generator, bringing significantly sharper audio and tighter lip sync to every generation. Dialogue now tracks mouth movements much more accurately while sound matches the scene context, helping creators generate clips that look and sound far more natural.
Researchers just introduced Vista4D, a framework that converts standard video into dynamic 4D point clouds. It lets users reshoot footage from any camera angle while keeping motion consistent, and even add or remove objects directly inside the 3D space. Why it matters: It turns static single-angle recordings into fully controllable virtual production sets.
Odyssey just introduced Odyssey-2 Max, a 3x-larger world model capable of continuous, real-time physical simulation. Unlike traditional video generators, it predicts open-ended futures autoregressively to internalize real-world mechanics, topping VBench 2 physics scores. It is currently available in private beta for robotics and gaming partners.
Researchers just introduced CoInteract, a video framework that turns a product photo and a person photo into a realistic demonstration clip. By using a dual-stream design to model physical contact alongside RGB pixels, it prevents hands from clipping through objects.
YouTube just expanded its likeness detection tool to talent agencies and celebrities, helping them track down and remove AI face-swaps. The system works similarly to Content ID, and the kicker is that entertainers can now use it even if they don't have a YouTube channel of their own.
TwelveLabs just unveiled Pegasus 1.5 and Rodeo at NAB Show 2026, alongside an integration with Autodesk Flow Capture. Pegasus 1.5 extracts time-based structured metadata in a single API call, outperforming Gemini 2.5 Pro by 30% on video segmentation benchmarks. Meanwhile, Rodeo acts as an AI co-pilot that lets editors find, cut, and assemble footage using plain language.
Lightricks just released an IC-LoRA for its LTX-2 video foundation model that upgrades generations from 8-bit SDR to 16-bit HDR. Weighing around 340MB, it plugs right into existing ComfyUI pipelines to dramatically boost dynamic range and color depth. Why it matters: creators can finally color grade AI video footage inside standard editors without clipping shadows or highlights.
Netflix is rolling out a TikTok-style vertical video feed across its apps this month to help users discover shows, movies, and video podcasts. The streaming giant is upgrading its recommendation engine with modern AI architectures while leaning into generative AI for film production, backed by its recent acquisition of Ben Affleck's AI startup InterPositive. Why it matters: Netflix isn't just fighting for the living room TV anymore—it wants your mobile scrolling time too.
HeyGen just open-sourced HyperFrames, an agent-native framework that converts HTML into MP4 videos, built entirely inside Claude Code. It allows AI agents to programmatically generate video workflows, and HeyGen even proved the concept by producing its own launch video inside Claude Code.
Browser Use just open-sourced video-use, allowing coding agents like Claude Code to cut videos directly. Instead of burning millions of tokens on video frames, it feeds the LLM word-level audio transcripts and on-demand timeline snapshots to trim dead air, strip filler words, and burn subtitles.
ByteDance and CUHK researchers just introduced OmniShow, an end-to-end video model built specifically for human-object interactions. It accepts product images, character photos, audio tracks, and pose skeletons all at once, generating up to 10-second marketing clips with synced lip-sync and hand movements. The code is currently under internal review on GitHub ahead of an open-source release.
Researchers just introduced LPM 1.0, an AI model that turns a single photo and audio input into an interactive, real-time visual avatar. Rather than simple lip-syncing, it captures natural idle behaviors, gaze shifts, and pauses while keeping the character's identity stable for up to 45 minutes. It offers a glimpse into true visual conversational partners, though it remains a closed research project for now.
Meta is reportedly training an AI avatar of CEO Mark Zuckerberg to interact with employees and hand out feedback in meetings. The digital twin learns from his voice, mannerisms, and public statements, with Zuckerberg himself spending up to 10 hours a week coding on AI projects. If it works, Meta plans to let everyday creators build their own meeting-ready clones next.
