AI Daily Brief — 11 June 2025

Hollywood took its first direct shot at a generative AI vendor. Anthropic killed its own AI marketing blog after transparency criticism. Mistral’s first reasoning models absorbed first-day reception. OpenAI rolled o3-pro into broader markets. ChatGPT recovered.

Top stories

• Disney, NBCUniversal and DreamWorks file 110-page copyright suit against Midjourney. US District Court LA. Calls Midjourney a ‘virtual vending machine’ for unauthorized copies of Shrek, Homer Simpson, Darth Vader and 150+ characters. First time major Hollywood studios have sued an AI company directly. Seeking $150K per infringed work. via CNN
• Anthropic shuts down ‘Claude Explains’ AI-generated blog. Less than a month after launch, Anthropic quietly wound down the marketing blog after backlash over transparency about how much was actually written by Claude. Site redirected to the Anthropic homepage. 24+ sites had already linked to it. via The AI Insider
• Mistral Magistral reasoning models — first-day coverage. Magistral Medium hit 73.6% on AIME2024 (90% maj@64); Small 70.7%/83.3%. Multilingual chain-of-thought; promoted at ~10x speed in Le Chat. On GPQA/AIME still trails Gemini 2.5 Pro and Claude Opus 4.
• ChatGPT / Sora outage stretches into Wednesday morning. Tuesday’s service disruption (elevated error rates across 21 ChatGPT components plus Sora) ran more than 10 hours, with full recovery confirmed Jun 11. Most severe blackout since March 2025.
• OpenAI o3-pro rolls out widely. Replaces o1-pro for ChatGPT Pro/Team and the API. Xinhua and global press cover. Enterprise/Education access promised the following week. Priced at $20/M input, $80/M output. via Xinhua
• Altman’s ‘The Gentle Singularity’ reverberates. ‘We are past the event horizon; the takeoff has started.’ Day of commentary across tech press and academia.
• Meta-Scale AI deal preview reporting. Detailed Zuckerberg’s push to invest several billion into Scale AI and recruit CEO Alexandr Wang to lead a new AGI push — precursor to the $14.3B deal formally announced Jun 12-13.

Who shipped

No fresh model launches. Disney + NBCUniversal + DreamWorks shipped legal action. Anthropic shipped a quiet retraction.

Quiet corners

No major release from Google DeepMind, xAI, DeepSeek or Qwen/Alibaba dated Jun 11. No notable arxiv splash specifically time-stamped Jun 11 (the Magistral paper formally posted Jun 12).

By the numbers

• 110 / 150+ — pages in Disney lawsuit / characters listed
• $150K per work — Hollywood’s statutory-damages claim
• $20 / $80 per million tokens — o3-pro pricing
• 10+ hours — ChatGPT outage total
• Most-mentioned company: Midjourney

Compiled by AI Feed’s editor from verified web sources for 11 June 2025.