The festival takes place at the Meliá Calviá Beach resort and is backed by the Ayuntamiento de Calvià, the local municipal authority, which is positioning Mallorca as a base for AI filmmaking in Europe. The jury includes Hollywood producer Tommy Harper among its members. Full jury lineup and submission information at aimovieawards.org. Partnership and hospitality packages are available.

More Information Here

More Information Here

More Information Here

OpenAI announced tonight that it is shutting down the Sora app and exiting AI video generation entirely. ChatGPT will also stop generating video from text prompts. The Disney deal, which included licensing more than 200 characters and a planned $1 billion investment, is dead with it. OpenAI says the Sora research team will refocus on robotics, but the real driver is cost: the company is trimming expensive consumer products ahead of an expected IPO. For anyone who built workflows around Sora, the priority is immediate. Get your videos and assets off the platform before shutdown timelines are confirmed. For the full story, see our breaking News coverage.

DeepBrain AI Moves From Static Avatars to Conversational Video Agents

DeepBrain AI, the Palo Alto-based avatar platform, has launched what it calls AI Video Agents for enterprise customers. The shift is from one-way video, where an AI avatar delivers a scripted message, to two-way interaction, where the avatar listens, responds and adapts in real time. The company says the technology is already deployed with SAP, Shinhan Bank and Samsung Securities across customer service, internal training and knowledge management. For anyone tracking where AI video meets actual business revenue, this is worth noting. The use case isn’t content creation. It’s operational: replacing help desks, onboarding sessions and customer support workflows with interactive video avatars that scale without headcount.

OpenAI Plans to Nearly Double Its Workforce by Year End

The Financial Times reports that OpenAI intends to expand from roughly 4,500 employees to around 8,000 by the end of 2026, hiring across product, engineering, research, sales and what it describes as technical ambassadorship. In light of tonight’s Sora shutdown, this hiring spree looks different. The growth in sales and enterprise functions now points squarely at text, code and business tools, not creative products. According to Ramp’s AI Index, businesses are now 70% more likely to choose Anthropic over OpenAI when purchasing AI tools for the first time. OpenAI’s response is to put people on the ground selling enterprise contracts for the products it’s keeping, not the ones it just killed. The message to the AI video industry is hard to miss: OpenAI has decided that video generation isn’t worth the compute.

The move follows ByteDance’s decision to suspend a broader global launch of Seedance 2.0 earlier this month after copyright pressure from Hollywood studios and a bipartisan intervention from US senators. Rather than wait for that situation to resolve, ByteDance appears to be taking a different route: embedding the model inside an existing app, in markets where US IP enforcement carries less immediate weight.

The CapCut integration puts Seedance 2.0 directly into the editing timeline. Users can access the model through AI Lab, AI Generator, and a new Video Studio canvas. Features include text-to-video generation for clips up to 15 seconds, multi-shot storytelling, built-in dialogue and lipsync, spatial sound, and reference support for maintaining character and style consistency across scenes. Notably, the initial rollout restricts the ability to generate videos from images or videos containing real faces, a direct response to the controversy that followed the model’s February launch in China, when viral clips featuring unauthorised celebrity likenesses triggered cease-and-desist letters from Disney and the Motion Picture Association.

For creators and editors already working in CapCut, particularly in the supported regions, this is the first opportunity to test Seedance 2.0 within a production workflow rather than through standalone demos. The practical advantage is workflow integration: generating a clip and editing it happen in the same application, which removes the export-import friction that slows down most AI video tools.

The strategic significance is harder to miss. CapCut already has hundreds of millions of users. Embedding a generative video model inside that user base, even in a limited rollout, creates adoption at a scale that standalone AI video tools can’t match. If the regional launch goes smoothly, ByteDance will have a strong case for expanding availability, potentially before the US copyright questions are resolved.

For UK-based creators and production teams, the wait continues. None of the initial rollout markets are in Europe or North America, and ByteDance has given no timeline for expansion beyond the seven launch territories. Given the unresolved copyright disputes with Hollywood studios and ongoing US Congressional scrutiny, Western availability is likely to depend on how smoothly the current rollout goes and whether ByteDance can demonstrate that its content safeguards hold up under real-world use. For now, the strongest AI video generation model on the market remains out of reach for most of this publication’s readers.