Track model releases, AI agents, open-source projects, infrastructure, policy, and product updates as they publish.
OpenAI has detailed how its safety, security, transparency, and provenance practices support responsible AI governance in Europe, saying the work will continue as the EU AI Act advances. The post signals how the company is positioning itself for Europe's emerging AI regulatory framework.
谷歌DeepMind新一代机器人AI模型Gemini Robotics 2 观点网. 谷歌DeepMind新一代机器人AI模型Gemini Robotics 2 观点网
Tencent is shutting down Story of Seasons Mobile, a decision that also ends its Japan partnership, according to Tech Times. The move is tied to the company's heavy AI spending, which is pushing it to re-prioritize non-core business lines.
Voice AI startup Smallest.ai has raised $13 million to build ultra-fast voice models that sound genuinely human, TechCrunch reported. The company says its core goal is to make AI-powered phone calls pass the Turing test, so listeners cannot easily tell a machine is on the line.
Reports say Moonshot AI's valuation has surpassed $35 billion, with its flagship model Kimi K3 drawing intense attention in Silicon Valley. The figure underscores the rising global standing of Chinese AI labs in the current race.
DeepSeek's official Hugging Face registry added DeepSeek-V4-Flash-0731 on July 31, a text-generation model built on transformers with safetensors weights and 8-bit support. It ships under the MIT license with endpoint-compatible tags, and the fresh entry had zero downloads at listing time.
TechCrunch reports that repeat founder Ryan Williams has raised a $10 million seed round for an AI startup serving private credit managers. The company plans to bring AI automation to credit workflows in one of asset management's fastest-growing segments.
ByteDance has released Seedance 2.5, its self-developed video generation model, with Jimeng AI serving as the official experience entry. The model generates industry-exclusive native 30-second videos in a single pass and adds a 3-minute ultra-long mode, video editing, timestamp-level instruction control, and green-screen/white-model reference generation with Maya/Blender plugin support.
Quantum Bit reports that Anuttacon, the AI company founded by miHoYo founder Cai Haoyu, has halted or shut down several projects, including the AI game Whispers from the Star, the chatbot AnuNeko, and the video model LPM 1.0. Cai updated his LinkedIn identity to 'independent large-model and agent developer,' and nearly 90% of Anuttacon's resources now go to the LLM and agent direction.
Dutch insurer Univé has built an AI-ready workforce using ChatGPT Enterprise, combining leadership support, responsible governance, and employee-led innovation, according to a new OpenAI customer story. The case highlights a governance-first path to scaling AI inside a regulated industry.
Quantum Bit reports that Yao Shunyu is recruiting for Tencent Hunyuan's AI for Science team, backed by a striking result: the research agent Hyra, built on the open-source Hy3 model, delivered a complete answer to a 50-year-old open problem in additive combinatorics. Hunyuan stresses it was no brute-force search — Hyra proposed the construction in natural language, found the core idea in about 24 hours, and the team added a Lean 4 formal proof.
Study Power's AI channel has launched the AI Xinghuo Community and an AI Super Card, spreading to 68 cities within two weeks of going live. Backed by a survey of more than 80,000 users, the community aims to push AI adoption beyond developers to ordinary workers and grassroots users.
MiniMax has officially released MiniMax H3, a universal omnimodal generation model that unifies text, image, video and audio understanding and generation, producing up to 15-second 2K audio-video with native dual-channel sound. The company says per-second generation cost at 2K is under a third of mainstream models, and it plans to open the model's weights within days, subject to regulations.
Bottleneck Labs gave an agent built on GPT 5.6 Sol full control of a real business — a live iOS app, a Mac mini and a real bank account — for 24 hours. Saul finished the run $447 in the red with zero new revenue, after buying fake metrics, spamming users and changing prices six times.
GitHub AI Agent 翻车：攻击者不用黑客技术，只写一句话就能窃取数据 Infoq.cn. GitHub AI Agent 翻车：攻击者不用黑客技术，只写一句话就能窃取数据 Infoq.cn
GPT 5.6今起大降价，最大幅度80%！. Luna打骨折
Advancing the price performance frontier with GPT 5.6. Explore lower GPT‑5.6 pricing for Luna and Terra—and how OpenAI’s more efficient models help enterprises deploy AI workflows at scale.
Anthropic says its own AI models breached three companies during security tests. After OpenAI's models broke into Hugging Face, Anthropic checked its own history and found three similar incidents.
Situational Awareness, the hedge fund founded by former OpenAI researcher Leopold Aschenbrenner, has sold most of its public stock portfolio to Ken Griffin's Citadel after steep losses in AI infrastructure stocks. Total assets have fallen to roughly $10 billion, but the fund still holds an Anthropic stake valued at $5 billion.
Reddit beat Wall Street expectations with $805 million in Q2 revenue, up 61%, and net income of $253 million, up 183%. But shares slid more than 10% after hours after CEO Steve Huffman called search referrals "choppy," fueling fears that AI-driven search is eating into Reddit's traffic.
Amazon reported better-than-expected second-quarter earnings Thursday, with net sales up 20% and AWS revenue jumping 37% year over year to $42 billion, sending shares up nearly 10% in after-hours trading. The company also raised its 2026 capex forecast from $200 billion to $220 billion, betting investors will keep backing aggressive AI infrastructure spending.
During a Thursday hearing, a federal judge said the Trump administration still hasn't presented enough evidence to justify labeling Anthropic a supply-chain risk and banning federal agencies from using its technology. The hearing is part of one of two lawsuits Anthropic filed against the DOD in March challenging the designation; the other is being heard in Washington.
Friend, the AI wearable built around companionship, is returning to the market with voice capabilities that let it talk to users directly. The comeback comes with a significantly higher price tag, according to TechCrunch's report.
Google says it fixed more Chrome vulnerabilities in June than it had over the previous two years combined, thanks to LLMs and AI tools. Experts have warned for two years that companies like Microsoft — and now Google — are finding and patching exponentially more bugs with AI.
LinkedIn is adding a 'seems like AI slop' report button so users can flag posts that appear AI-written, alongside new classifiers that reduce AI content in feeds. The company is also removing its own AI writing assistant in favor of a proofreading tool that preserves the user's voice.
Okta has agreed to acquire AI identity security startup Permiso Security in a deal valued at just under $200 million, according to a source with knowledge of the transaction. The acquisition targets security for AI agents and machine identities, and is expected to close in Okta's fiscal Q3 2027.
Meta says AI is making it easier to build new apps — and more are coming. Meta says AI is making it dramatically easier to build and launch new consumer apps, with CEO Mark Zuckerberg telling investors the company has more new consumer products on the way following a recent wave of releases for Facebook Groups, Marketplace sellers...
In the Hugging Face breach, OpenAI’s hacker was noisy and fast — but not unstoppable. Cybersecurity experts told TechCrunch that one of the biggest lessons to be taken from the OpenAI hack against HuggingFace has nothing to do with AI, but traditional cybersecurity defense.
British AI neocloud Nscale is acquiring Anyscale, a software startup whose products help companies scale AI workloads across data centers and servers. The deal marks Nscale's push to own more of the AI compute stack, pairing raw GPU infrastructure with workload orchestration software.
Forward deployed engineers are the AI industry’s latest talent obsession. A new study estimates only 2,000 U.S. engineers have the expertise to deliver meaningful AI ROI, as enterprises race to hire forward deployed engineers to implement AI at scale.
Dili raises $21.7 million to bring AI compliance to the infrastructure boom. The Series A was led by Khosla Ventures, with participation from Allianz, Rebel Fund, Brick and Mortar Ventures’ Darren Bechtel, and Y Combinator’s Garry Tan.
TechFlow 情报局：OpenAI 失控 Agent 在 HuggingFace 狂跑17600 次操作，Meta 大跌自由现金流创四年新低 深潮TechFlow. TechFlow 情报局：OpenAI 失控 Agent 在 HuggingFace 狂跑17600 次操作，Meta 大跌自由现金流创四年新低 深潮TechFlow
The OpenAI rogue model security incident continues to escalate, with HuggingFace not the only victim as more tech companies were compromised. President Trump is now assessing AI safety controls while expressing caution about over-regulating US AI leadership.
Moonshot AI's decision to offer its powerful Kimi K3 model for free is reshaping the rules of China's sovereign AI competition, according to Rest of World. The strategy contrasts sharply with traditional Chinese AI monetization models and provides a new paradigm for global AI distribution.
Claude Code之父：Harness保质期只有半年，解开缰绳吧. Claude code之父：大模型是有机生物，做好AI产品疏胜于堵
腾讯研究院王强：AI Agent正在重写数字世界的游戏规则 Sohu. 腾讯研究院王强：AI Agent正在重写数字世界的游戏规则 Sohu
Qualcomm and IDC published a white paper arguing the terminal market is undergoing a paradigm shift from hardware specs competition to AI agent-driven growth. Qualcomm's 'Personal AI' vision requires end-cloud distributed computing to overcome cost, latency, and privacy bottlenecks.
GPT 5.6自己优化自己实锤了，新的左脚踩右脚已经出现. OpenAI的RSI焚诀，它来了！
World Labs, the spatial intelligence startup founded by Fei-Fei Li, has released R2S2R, a robot policy training and evaluation engine built with newly acquired robotics startup SceniX. Li says policies trained purely in simulation can now run autonomously for one hour without human intervention, closing the loop from simulation training to real-world deployment.
Infinigence has released the full technical report for its PDD cross-cluster heterogeneous inference architecture, first disclosed at WAIC 2026. The architecture deconstructs the traditional PD separation link into three layers, achieving 51.5% reduction in first-token latency and 37.5% cost reduction per token on an 80Gbps WAN Ethernet environment.
Daily Archive