AI Video Generation — Latest News and Analysis

YouTube has launched AI avatars for Shorts, letting creators generate deepfake clones of themselves using face and voice capture powered by Google's Veo models.

Three YouTube creators have filed a class action DMCA lawsuit against Apple for allegedly scraping videos to train AI models, setting a major legal precedent.

Google has upgraded Vids with Veo 3.1 video generation, Lyria 3 music creation, and directable AI avatars, offering 10 free clips per month to all users.

Google has launched Veo 3.1 Lite, a low-cost AI video model for developers, as OpenAI exits the market after shutting down Sora over massive losses.

Midjourney has revealed its annual revenue significantly surpassed $200 million, with the profitable AI image generator now eyeing hardware expansion.

ByteDance has launched its Dreamina Seedance 2.0 AI video model in CapCut after adding copyright safeguards following Hollywood backlash over actor likenesses.

OpenAI has shut down its Sora video-generation app, killing a $1 billion Disney deal and retreating from consumer video as it refocuses ahead of its IPO.

OpenAI has reportedly planned to fold its Sora video generator into ChatGPT after standalone app installs fell 45 percent in January 2026, raising cost and safety concerns.

OpenAI has been found testing a $100/month ChatGPT Pro Lite plan in checkout code, offering unlimited top-model access and 3-5x reasoning quotas over Plus.

ByteDance has launched a hiring push for nearly 100 US AI roles in its Seed division amid Hollywood copyright disputes over its Seedance 2.0 video generator.

A federal court has ordered OpenAI to stop using 'Cameo' for its Sora 2 video feature, ruling the name infringes on the celebrity platform's trademark.

ByteDance has pledged to strengthen safeguards for its Seedance AI video tool after Disney threatened legal action over copyright infringement.

Meta has expanded Vibes into a standalone app for AI-generated videos, testing in Brazil and Mexico as it positions itself against OpenAI's Sora.

OpenAI's Sora video app has suffered a 45% download collapse in January, falling from No. 1 to No. 101 on the US App Store just three months after launch.

Google has made Project Genie available to AI Ultra subscribers in the US, enabling AI-generated playable 3D worlds with 60-second session limits.

Research from NewsGuard found ChatGPT missed 92.5% of AI-generated Sora videos in testing, while Grok failed 95% and Gemini achieved a 78% error rate, exposing critical limitations in AI content verification.

Adobe has released Premiere Pro 26.0 with AI-powered Object Mask that generates precise masks with one click and delivers 20x faster tracking performance.

The updates make videos more expressive and creative with simple prompts, add native vertical video support in 9:16 aspect ratio, and provide advanced upscaling.

MiniMax has raised $619 million in its Hong Kong IPO at a $6.5 billion valuation, joining Zhipu AI as Chinese firms beat OpenAI to public markets.

New data has revealed that 20% of YouTube recommendations are now "AI slop," generating millions in revenue.

YouTube has permanently banned popular AI trailer channels Screen Culture and KH Studio, citing spam policies just days after Disney issued a cease-and-desist to Google.

Google has purged dozens of AI-generated Disney videos from YouTube, complying with a cease-and-desist order just days after Disney's $1B OpenAI deal.

Disney has invested $1 billion in OpenAI to license IP for Sora while simultaneously suing Google for copyright infringement on a massive scale.

McDonald's has removed its AI-generated Christmas ad following viewer backlash over "uncanny" visuals, highlighting the risks of using early-gen AI tools.

Kling AI has launched Video O1, a unified multimodal model that merges video generation and editing, claiming superiority over Google Veo 3.1.

Apple Research has unveiled STARFlow-V, a 7-billion parameter model using Normalizing Flows to solve error accumulation in generative video.

Runway has released Gen-4.5, a new AI video model that has claimed the top spot on independent benchmarks, outperforming Google's Veo 3 and OpenAI's Sora 2 Pro.

ByteDance has released Vidi2, a 12B open-source AI model featuring Spatio-Temporal Grounding for precise video editing, challenging GPT-5 and Gemini 3.

Tencent has released HunyuanVideo-1.5, an 8.3B parameter open-source model optimized to run on consumer GPUs with 14GB VRAM.

ElevenLabs has expanded its Studio platform to include image and video generation, integrating OpenAI Sora 2, Google Veo 3.1, and Kling 2.5 into a unified timeline editor.