In partnership with
Good Morning,
Google AI Overviews grow. Nvidia backs Safe Superintelligence with Vera Rubin access and a reported $5 billion investment.
Nvidia and Microsoft launch the Open Secure AI Alliance without OpenAI, Google, or Anthropic, as Microsoft unveils MAI-Cyber-1-Flash and Project Perception for agentic cybersecurity.
Moonshot releases Kimi K3 weights, Anthropic rejects open-weight bans while calling for tougher safeguards, OpenAI’s Hugging Face breach revives security-versus-alignment concerns, and AI product risks spread across Claude, Google Search, and Meta’s app ecosystem.
Here’s your daily update.
Markets
The Nasdaq fell 0.18%, led by semiconductors, as advances by Chinese chipmakers intensified competition concerns across the AI supply chain. Nvidia dropped 5.0%, while SanDisk and Micron lost 11.0% and 2.3%. Microsoft, Alphabet and Apple bucked the weakness, gaining between 1.2% and 2.1% ahead of major tech earnings.
Headlines
Nvidia and Safe Superintelligence announce a long-term partnership that gives Ilya Sutskever’s lab access to Vera Rubin systems and a reported $5 billion Nvidia investment. (Reuters)
Nvidia and Microsoft launch the Open Secure AI Alliance with dozens of partners, notably without OpenAI, Google or Anthropic. (The Verge)
Microsoft unveils MAI-Cyber-1-Flash and Project Perception, its first major AI cybersecurity model-and-agent platform. (Axios)
Moonshot AI releases Kimi K3 weights and a technical report, making its 2.8T-parameter open-weight model available to developers. (Hugging Face)
Anthropic says it does not support banning open-weight models, while warning that Chinese AI models still need tougher testing and safeguards. (TechCrunch)
OpenAI sees its Hugging Face breach reignite debate over whether frontier-model failures are primarily cybersecurity problems or deeper alignment problems. (TechCrunch)
Anthropic faces privacy scrutiny after some shared Claude chats and Artifacts appear to have been indexed by Google. (TechCrunch)
Google sees AI Overviews become increasingly default in search, now appearing in 43% of searches according to new Similarweb data. (TechCrunch)
Meta rolls out Meta AI inside Threads direct messages as it pushes its chatbot deeper across its app ecosystem. (TechCrunch)
Speak naturally. Send without fixing.
Wispr Flow turns your voice into clean, professional text you can send the moment you stop talking. Not rough transcription you have to clean up. Actual polished text — ready for email, Slack, or any app.
Speak the way you think. Go on tangents. Change your mind mid-sentence. Flow strips the filler, fixes the grammar, and gives you text that reads like you spent five minutes writing it.
89% of messages sent with zero edits. Millions of professionals use Flow daily, including teams at OpenAI, Vercel, and Clay. Works on Mac, Windows, and iPhone.