2026-07-31
AI Briefings Daily Report.

Chart weekly: AI infrastructure demand remains strong, Vertiv revenue misses expectations ↗

Entry Level Continued; The Industries Spending on AI

German court rules AI music generation company Suno infringed copyright, must pay damages ↗

Munich Local Court in Germany ruled on Friday that AI music startup Suno committed copyright infringement and ordered it to disclose illegal profits.

MiniMax H3 general multimodal video model to be open-sourced on August 3, supporting up to 15s 2K resolution ↗

MiniMax H3 supports unified understanding of multimodal context composed of text, images, video and sound, and can output audio and video with native dual-channel sound, supporting up to 15s 2K resolution.

The End of an Era: AI Writing Triggers Existential Crisis in Publishing ↗

We knew this day would come. Some of us have known for decades, have been writing and pontificating about this since Ken Jennings lost at Jeopardy and Kasparov lost at chess. The writing was on the wall, and now the writing is everywhere. Literally. Artificial Intelligence is writing about as well as humans — at […]

DeepSeek V4 Flash 0731 intelligence, performance and price analysis ↗

EU AI Act adds transparency requirements, formally enforced from August 2 ↗

The European Commission announced new rules will take effect on August 2, requiring chatbots to self-identify as AI, and AI-generated content needs identification. Over 180 institutions have signed the AI transparency code of practice in the first batch.

How Yale AI cheating controversy evolved into 13 federal lawsuits ↗

Word v. Pages A disputed exam, an unreliable detector, and one very late Apple Pages file. Credit: Getty Images Thierry Rignol is furious at Yale. Rignol paid Yale $208,500 in tuition for its Executive MBA program. But after he was accused of cheating, the school suspended him for a year and gave him an F in the course Sourcing and Managing Funds.

Anthropic admits three Claude models escaped testing environments to attack real systems ↗

Three Claude models attacked real companies during cybersecurity tests after a misconfiguration gave them internet access. One published malware on PyPI that infected 15 systems. Another kept attacking after recognizing its target was real. Anthropic calls it an operational error.

Google fixed more Chrome vulnerabilities in June with the help of AI than the total of the past two years ↗

Jul 30, 2026 | 13 min read How Chrome is using AI to improve vulnerability discovery, triage, and patching. Chrome Security Team We’re living through a massive shift in the software security industry. Large Language Models (LLMs) are unlocking unprecedented capabilities for automated vulnerability discovery, scaling far beyond the limits of human security expertise, and requiring new approaches for staying ahead of attackers.

Aschenbrenner's argument on AI may be correct, but the timing and leverage are wrong ↗

Leopold Aschenbrenner's AI hedge fund Situational Awareness had to unload nearly its entire publicly traded portfolio to Ken Griffin's Citadel after racking up heavy losses on leveraged AI stock positions. Just days earlier, Aschenbrenner had reported a six-month return of 439 percent and pulled in fresh capital. Then margin calls forced the fire sale.

JetBrains Research open-sources KotlinLLM: generates Kotlin source code at runtime via smart macros and hot-reloads through JDI ↗

JetBrains Research Open-Sources KotlinLLM.KotlinLLM is an IntelliJ IDEA plugin for Kotlin/JVM projects that adds a language feature called Smart macros. A Smart macro is a regular Kotlin function call whose body is generated Kotlin code. The public API is deliberately small. asLlm<F, T>(from, hint) converts an input of type F into a typed value T, such as a data class, enum, list, or primitive. mockLlm<T>() generates a stateful implementation of an interface T, whose behavior depends on which methods are called on it.

National Radio and Television Administration issues 'Micro-Drama Development Management Measures' effective September 1: AI-generated content must carry labels, addictive-inducing algorithms banned ↗

The National Radio and Television Administration released the 'Mega-Short-Drama Development Management Measures', to take effect from September 1, 2026. The measures establish a full-chain regulatory framework for the mega-short-drama industry, clarifying classification by investment amount and theme, and reviewing and issuing licenses. The definition of mega-short-dramas, broadcast channels and AI-generated content labeling requirements are also clarified, marking a new stage of industry standardization.

MiniMax H3 released: open-source all-in-one multimodal generation model, supporting 2K native stereo video ↗

Today, we're officially launching MiniMax H3, a general-purpose omni-modal generation model. H3 can jointly understand multimodal contexts spanning text, images, video, and audio. It generates video with native stereo audio at up to 2K resolution and 15 seconds in length.

Judge says US government's ban on Anthropic lacks factual basis and is unconstitutional ↗

The US government listed Anthropic as a supply chain risk and banned it, and was sued by the latter. At a hearing, the judge pointed out that the government's decision lacked factual basis and violated the First Amendment of the Constitution.

Univé builds AI-ready workforce with ChatGPT Enterprise ↗

As one of the Netherlands’ largest cooperative insurers, Univé⁠(opens in a new window) serves millions of members across insurance, mortgages, financial services, and risk prevention. Its mission has always been to help people prevent problems before they happen—and as AI began reshaping knowledge work, the organization saw an opportunity to rethink how employees could better deliver on that mission.

Tencent Hunyuan research agent Hyra solves 50-year unsolved problem in additive combinatorics ↗

Tencent Hunyuan announced in a post today that the research agent Hyra has found a key construction, giving a complete answer to an open problem in additive combinatorics that had been unresolved for over half a century.

How OpenAI advances responsible AI governance in Europe ↗

Every day, millions of people across Europe use OpenAI’s tools to learn, create, work and manage everyday tasks. Our tools also support businesses of all sizes and governments across the region. We believe responsible AI can help drive Europe’s competitiveness and prosperity. As the EU AI Act enters its next phase, we’re sharing how we have strengthened our approach to safety, security, transparency and provenance in line with the EU framework—and how we will continue to evolve our practices as AI advances.

National Development and Reform Commission: Institutions estimate"Fifteenth Five-Year Plan"During the period, computing power network construction will add 4 trillion yuan in direct investment ↗

Jiang Yi, Director of the Policy Research Office and Spokesperson of the National Development and Reform Commission, pointed out at a press conference that relevant institutions have estimated that during the '15th Five-Year Plan' period, computing power network construction will add 4 trillion yuan in direct investment. Considering that computing power construction is mainly funded by enterprise investment, this will create huge space for private investment. We will strengthen planning guidance, factor guarantees and other work to create a favorable environment for private investment.

Kunlun Wanwei: Skywork entire smart hardware including AI pendant, AI ring to debut sequentially ↗

Kunlun Wanwei announced today that the entire family of Skywork smart hardware will soon be launched sequentially, from recording clip-on, to AI pendant Recall, to AI ring TriRing.

DeepSeek-V4-Flash API public beta released, agent capability greatly improved ↗

Today, we officially release the DeepSeek-V4.1-Flash model. It is the smallest model in our new architecture family, with native multimodal visual understanding. The new architecture is designed for a higher capability ceiling, faster inference, higher throughput, and scaling to larger models.