AI daily

2026-08-31

EU Classifies ChatGPT as Very Large Search Engine; MiniMax Unveils Real-Time Video Model

AI DAILY BRIEFING

The EU Commission has officially classified ChatGPT as a very large search engine under the Digital Services Act, citing at least 45 million monthly EU users. This imposes new obligations on OpenAI, including risk assessments, transparency reports, and an ad archive by end of 2026. Legal experts debate whether the Commission can demand access to training data. Meanwhile, MiniMax and fal released H3 Max, a video generation model that produces 5-second videos in under 3 seconds, enabling real-time live streaming and AI short-video apps. The model tops image-to-video leaderboards but is limited to 768p resolution.

EU regulatory scrutiny on AI models is intensifying, with ChatGPT now facing DSA obligations similar to large platforms, potentially setting a precedent for other AI services.

Real-time video generation is becoming a reality, with H3 Max's speed enabling new applications like live streaming, though resolution limits may constrain some use cases.

The AI industry is witnessing a shift from model-centric to harness-centric competition, as seen in coding tools, and a growing demand for Macs as AI infrastructure.

Watch next: Watch for OpenAI's response to EU obligations and whether other AI models will face similar classifications. Also monitor the adoption of real-time video generation in consumer apps and the evolution of AI coding harnesses.

Featured

SSignal88
量子位
1 sources

MiniMax Launches H3 Max: Real-Time Video Generation Faster Than Playback

MiniMax and fal have jointly released the video generation model H3 Max, optimized for real-time generation, producing a 5-second video in under 3 seconds, faster than playback speed. The model has enabled real-time live streaming and AI short-video apps, with three developers independently creating AI live streams within a week. H3 Max ranks first in image-to-video on multiple leaderboards, though resolution is limited to 768p.

SynthePulse InsightRead deep analysis
APriority78
The AI Insider
1 sources

Caterpillar Leverages Decades of Automation Experience to Expand AI Deployment

Caterpillar is leveraging its long history in autonomous mining to broaden AI use across construction, manufacturing, and enterprise operations. The company has introduced the Cat AI Assistant for field technicians and is applying AI to digital twins and legacy software modernization. It plans to invest $100 million over five years to retrain its 118,000 employees in AI, autonomy, and robotics.

SynthePulse InsightRead deep analysis
APriority78
InfoQ AI/ML/Data Eng
1 sources

DoorDash's Flux Runs 130,000 Engineering Tasks Through Cloud-Based Agents

DoorDash has moved its engineering agent workloads from developer laptops to its Flux cloud platform. The platform automated 130,000 engineering tasks in one month and supports more than 25,000 automated code reviews weekly. Flux uses isolated Firecracker microVMs, an MCP gateway, reusable playbooks, and multiple invocation surfaces to run agent workflows with scoped access and centralized auditing.

SynthePulse InsightRead deep analysis

More

APriority78
机器之心
1 sources

HOMIE Gen2 Released! Unlocking the Scaling Law of Human Experience

Ropedia has released the new multimodal capture system HOMIE Gen2, designed to scale human experience into training data for robots. The company proposes a 'Human Experience Scaling Law,' suggesting that robot capabilities will improve with the scalable growth of high-quality human experience. HOMIE Gen2 features immersive capture capabilities and synchronized multi-modal recording, providing training-grade data for embodied AI.

SynthePulse InsightRead deep analysis
APriority76
OpenAI Blog
1 sources

A Milestone in Expanding Access to AI

OpenAI announced that ChatGPT Ads has reached a $1 billion annualized revenue run rate and is expanding globally. This milestone highlights progress in AI commercialization while supporting broader access to AI through free and affordable options.

SynthePulse InsightRead deep analysis
APriority79
机器之心
1 sources

OpenClaw 2.0 Released: Installation Overhaul, Browser UI Redesign, and Shared Cloud Sessions

OpenClaw 2.0 has been released, marking the project's largest update ever, with major changes including a rewritten installation process, a chat-first browser UI, migration to SQLite storage, and the introduction of shared cloud sessions. The team originally planned only two tasks, but the scope expanded into a major version, with the official blog joking it arrived faster than GTA 6. The update involved 933 contributors and over 16,000 merged pull requests, reflecting strong community engagement.

SynthePulse InsightRead deep analysis
APriority79
量子位
1 sources

Zhongqing Robotics CEO: 60% of resources to brain R&D to avoid becoming 'second Unitree'

Zhongqing Robotics founder Zhao Tongyang said after four months of seclusion that the company will allocate 60% of its capital and personnel to brain R&D this year, aiming to differentiate from Unitree. He has hired Li Liyun, former head of XPeng's autonomous driving, as CTO to improve brain capabilities from 60 to 80 points. This reflects intensifying competition in embodied AI, with companies shifting focus from hardware to AI brains.

SynthePulse InsightRead deep analysis
APriority76
InfoQ AI/ML/Data Eng
1 sources

Foundry Model Router Expands from Two Regions to 28, Refreshing Its Model Pool

Microsoft expanded Foundry's model router from two regions to 28 for global standard and 21 for data zone deployments, while adding Claude Opus 4.8 and GPT-5.6 and removing four deprecated models. Default deployments receive pool changes automatically; configured subsets exclude new models until added. The effective context window equals the smallest model in the pool.

SynthePulse InsightRead deep analysis
Selected
8
Sources
5
Featured
3
More
5
Previous2026-08-30Next2026-09-01
All reports