Daily AI News – 5 Minute Briefing on AI Trends (11 Sep 2025)

Welcome to Daily AI News. This column provides daily updates on the latest developments in the field of AI. We focus on sharing news, research, and innovative applications that are relevant to developers and those interested in the technology. Our goal is to help readers keep track of current trends and better understand practical uses and progress in AI-related areas.

For past AI daily news, please view Daily Tags.

1. OpenAI Signs $300 Billion Cloud Deal with Oracle

OpenAI has signed a roughly five-year deal with Oracle to purchase $300 billion worth of computing resources for AI data center construction. The deal requires 4.5 gigawatts of power capacity, equivalent to the output of two Hoover Dams, and the contract will take effect in 2027. This is one of the largest cloud service contracts in history, highlighting the AI infrastructure investment boom, despite OpenAI’s current revenue being far below this scale. Oracle has not immediately responded to details, but the deal will significantly boost its cloud business.

2. Thinking Machines Lab Releases First Research Blog, Aiming to Enhance AI Model Response Consistency

Mira Murati’s former OpenAI team, Thinking Machines Lab, launches the research blog “Connectionism,” with the first post exploring the defeat of non-determinism issues in LLM reasoning. Author Horace He points out that GPU kernel splicing introduces randomness, and reproducible responses can be achieved through precise control. This technology not only enhances reliability for enterprises and scientific research but also optimizes reinforcement learning training. The lab secured $2 billion in seed funding with a $12 billion valuation, and its first product is set to launch soon, focusing on custom AI models. This initiative challenges the AI community’s default acceptance of non-deterministic systems, driving more stable AI development.

3. Tech Giants Back RSL Standard to Prevent AI from “Ruining” the Internet

Companies like Reddit and Ziff Davis support the “Real Simple Licensing (RSL) Standard,” allowing creators to set licenses and prices for AI use of content. This non-profit standard, initiated by the RSS co-founder, has alliance support from Yahoo, Medium, and others. RSL aims to address AI’s free scraping of web content, collaborating with Cloudflare’s AI crawler blocking system to provide pricing control. While AI companies acknowledge the need for licensing, they have not yet complied with RSL.

4. YouTube’s Multi-Language Audio Feature Officially Opens to All Creators, Using AI for Video Dubbing

YouTube’s multi-language audio feature ends its two-year pilot and opens to millions of creators, using Google Gemini AI for automatic dubbing while preserving the creator’s tone and emotion. Tests show that non-native viewing time accounts for over 25%, such as the Jamie Oliver channel seeing viewership triple. The feature covers Facebook, Instagram, and Threads, supporting custom thumbnails to match audience languages. Pilot participants include MrBeast and Mark Rober; this expansion aids global content distribution, with a full rollout expected in weeks, promoting AI applications in content localization.

5. Spotify Upset Over Users Selling Data to Train AI Tools

Spotify warned users to stop selling Wrapped data for AI training, with about 10,000 users selling data via the Vana platform to generate personalized insights. Spotify issued cease-and-desist orders, but developers claim they haven’t received them. The incident sparks data privacy debates, emphasizing the need for user consent in AI training.

6. Ted Cruz Proposal: AI Companies Can Self-Regulate for Up to 10 Years

Senator Ted Cruz introduced the SANDBOX bill, allowing AI companies to apply for exemptions from federal regulations for up to 10 years to test innovations. The bill requires disclosure of risk mitigation plans to promote U.S. AI leadership, but raises concerns over regulatory laxity. The proposal is similar to a failed 10-year AI ban.

7. Google Removes Daily Hub AI Feature from Pixel 10

Google has removed the Daily Hub AI preview feature from Pixel 10 devices, which originally displayed daily summaries in Discover and home screen widgets. Google stated it will improve and relaunch it, as the current version has limited functionality. This removal highlights challenges in iterating AI features, with insufficient user feedback.

References

1. Oracle, OpenAI Sign $300 Billion Cloud Deal The Wall Street Journal. 2025-09-10
2. Thinking Machines Lab wants to make AI models more consistent TechCrunch. 2025-09-10
3. Thinking Machines X (Twitter). 2025-08-29.
4. Reddit, Ziff Davis Back New Idea to Stop AI From Ruining the Internet PCMag. 2025-09-10
5. YouTube’s multi-language audio feature for dubbing videos rolls out to all creators TechCrunch. 2025-09-10
6. Spotify peeved after 10,000 users sold data to build AI tools Ars Technica. 2025-09-10
7. Ted Cruz’s new bill would let AI companies set their own rules for up to 10 years The Verge. 2025-09-10
8. One of Google’s new Pixel 10 AI features has already been removed Ars Technica. 2025-09-10

Disclaimer: This daily news is for reference purposes only. For the latest, official information in AI, please follow authoritative channels.