Daily Highlights
- OpenAI closes $11B first tranche of new round at $1.25T valuation
- OpenAI launches GPT-6 Cyber for vetted security customers
- Anthropic IPO may slip; valuation could fall to ~$1T
- Thinking Machines nears $5B raise at $40B valuation
- Kuaishou's Kling AI completes ~$3B funding round
Today's Headlines
[OpenAI Closes ~$11B First Tranche of New Funding Round at $1.25T Valuation] According to people familiar with the matter, OpenAI has completed the first closing of roughly $11 billion in its new funding round, with the funds landing in its accounts as of Friday, bringing the post-money valuation to $1.25 trillion. Amid mounting concern over AI capex, both the round size and valuation came in below initial targets; the total may reach $20 billion as remaining tranches close in the coming weeks. The raise is a key funding move after OpenAI delayed its IPO. (Source: Sina Finance)
[OpenAI Launches GPT-6 Cyber, Restricted to Vetted Security Customers] This week OpenAI released GPT-6 Cyber, a model for cybersecurity professionals capable of finding previously unknown vulnerabilities and developing working exploits. OpenAI says the model uncovered 81 high-severity Firefox vulnerabilities and 948 Linux exploits, with access tightly restricted to prevent misuse. It marks the first large-scale, controlled release of offensive-capable frontier AI to defenders. (Source: Distill Intelligence)
New Products & Tech
[Planet Launches Planet AI: Query a Decade of Satellite Imagery With One Sentence] On Nov 5, US satellite data company Planet launched the Planet AI app, combining large language models with more than ten years of global satellite imagery archives. Users can search locations, track changes over time, and generate comparison or time-lapse imagery in natural language. Every AI conclusion is grounded in viewable raw imagery to reduce hallucination risk; the app is free to try at ai.planet.com. (Source: IT Home)
[Anthropic Rolls Out Claude Personal Finance in US Beta] On Nov 5, Anthropic launched Personal Finance in Claude, opening a US beta. The feature positions Claude as a personal financial planning assistant, marking another consumer-side productization effort by the company. (Source: Anthropic)
IPO & Fundings
[Thinking Machines Nears $5B Round at $40B Valuation] Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, is closing in on a new $5 billion funding round at a $40 billion valuation, according to multiple reports. The company also released its new Inkling 2 Small model on Nov 5, advancing on both capital and product fronts. (Source: Distill Intelligence)
[Kuaishou's Kling AI Completes ~$2.98B Round, a Record for Video AI] Kuaishou's Kling AI unit has completed a funding round of roughly $2.98 billion, the largest ever for a video-generation model company. The capital will fund continued model iteration and global commercialization; Kling previously reported over 100 million global users and ARR approaching $500 million. (Source: Distill Intelligence)
Open Source & Research
[Inkling 2 Small: a 594B-Parameter MoE With 17B Active, Open Weights Planned] On Nov 5, Thinking Machines introduced Inkling 2 Small, a mixture-of-experts model with 594B total and 17B active parameters, with open weights planned. It would be the company's first open release, seen as a direct answer to the Llama and Qwen open-weight camps. (Source: MarkTechPost)
[DeepSeek Releases Prototype Autonomous Coding Agent] According to Bloomberg, DeepSeek released a prototype autonomous coding agent able to carry out coding tasks independently over longer horizons, extending its open-source momentum in code models. The release narrows the agent-capability gap between open and closed frontier labs. (Source: Bloomberg)
[Anthropic Research: Claude Agents Can Get Smarter by "Dreaming"] On Nov 6, Anthropic published "The components of agent memory," organizing agent memory into four layers — session logs, memory notes, knowledge representation, and injected context — and introducing a "dreaming" mechanism in which the model reviews logs, corrects errors, and distills rules during idle time, significantly improving long-horizon task performance. (Source: IT Home)
Company & People
[Anthropic IPO Could Slip, Valuation May Fall to ~$1 Trillion] Reports including Benzinga say Anthropic's IPO could be delayed and value the company around $1 trillion, below its $1.5 trillion target, amid a tech selloff; its amended filing showed roughly $895 billion in compute commitments. Separately, OpenAI President Greg Brockman said OpenAI could be ready to go public in 2027. (Source: Benzinga)
[Mistral Acquires SkipLabs; Meta Buys AI Coding Startup] Mistral announced the acquisition of code-tool maker SkipLabs this week, while Axios exclusively reported that Meta has bought an AI coding startup, joining the race to own coding. M&A around coding tools is becoming the densest consolidation zone in the AI application layer. (Source: Axios)
[Judge Rejects xAI Bid to Dismiss NAACP Pollution Lawsuit] A judge rejected Elon Musk's xAI bid to dismiss an NAACP pollution lawsuit over emissions from its Memphis supercomputer site, meaning the case will proceed to substantive proceedings. (Source: TechCrunch)
--- Powered by AOYII.com
