
Anthropic targets IPO before Thanksgiving
The company is eyeing a public listing with a valuation between $1.25 and $1.75 trillion.
ddNews Β· a daily AI briefing for Dirk (generative video/ddStudio, Lit

The company is eyeing a public listing with a valuation between $1.25 and $1.75 trillion.

New tests show GPT-6 successfully breaching 80.8% of simulated corporate networks.

Nous Research retrained Alibaba's Qwen model to remove pro-Chinese propaganda while maintaining capabilities.

Tiger Data releases an open-source extension using PRISM indexing for scalable vector search.

The funding round marks a massive valuation jump, though it now requires private investor capital.
The company raised $11B in a new funding round.

Unveiling a new quantum error correction platform with Oak Ridge.

Reportedly raising an additional $11B at a $1.25 trillion valuation.

The decision to end the polarizing pickup marks a strategic shift in Tesla's EV lineup.
A new LLM designed specifically to automate AI training tasks.

Ricursive Intelligence releases a 65B parameter model with top-tier agentic coding performance on TerminalBench 2.1.

A new open-weight model from Ricursive Intelligence optimized for agentic programming.

The company raised another $11B in a new funding round, cementing its position as one of the world's most valuable entities.

New collaborative environments allowing users to build app-like AI experiences.

DeepSeek and GLM show hacking capabilities nearly on par with OpenAI's private cyber models.

Reports suggest China recruited an AI executive to spy on the NVIDIA CEO.

Democratic victories and ballot measures may reshape data center development.

reports indicate Anthropic is using its own models to build newer ones, doubling development speed.

Permanent AI agents running on virtual machines for complex, long-term projects.

Users can now build customized, app-like AI experiences (e.g., shopping hubs) within Muse.
Certified AI coworkers for enterprise IT and security teams.

Extremely compact model capable of generating both speech and music.

Mira Murati's Thinking Machines debuts a 594B MoE open-weight model rivaling Gemini 3.8 Flash.

Historic Democratic victories include 41 data center-related ballot measures, mostly opposing expansion.

Integrates agentic development and governance directly into the DevOps platform via MCP server.

Restores the Compiler API and provides first-class support for Vue and Svelte.

a GA service that proves exploitability in sandboxes and suggests PR fixes.

Adds quantum-ready TLS and support for external HSM/KMS keys.

In talks to acquire AI agent security startup Clutch for $500M.

A new security model isolating credentials from AI agent workloads via hypervisor enforcement.
A $350M deal to launch DBX Agent Cloud, a storage platform for AI model and agent developers.
Shares jumped 14% as revenue growth accelerated to 39%, driven by AI connectivity.
A specialized Astra model trained for authorized vulnerability research and exploit patching.

Planned for H1 2027 to handle planning, comparison, and booking.
Enterprises are prioritizing storage as they hit GPU utilization ceilings in AI stacks.

OpenAI releases a specialized model for vulnerability research, exploit development, and patching.

US users can now video chat with customizable animated AI avatars.

Stepping back from NDAs for data center projects following US midterm backlash.

political opposition to infrastructure could hinder future AI expansion.

Now supports 15-second video clips with conversational step-by-step editing.

Greg Brockman indicates the company is scaling enterprise market share to prepare for public listing.

An open-source 280B MoE model with only 14B active parameters per token.

Preparing infrastructure to support AI systems reaching 10 gigawatts.

Google updates its fast, cost-effective model with support for longer video/audio clips.
New focus on moving AI agents from demo mode to production via improved data architecture.

Using AI agents to formally verify the Stacks Project to eliminate "AI slop" in mathematics.

New "real-world" benchmark shows most AI agents failing complex tasks like chip design (under 12% success).
Focuses on placing databases directly within existing AI pipelines to reduce agent latency.

Public preview of streaming support for real-time AI-powered applications.

A $350M deal to launch DBX Agent Cloud for AI model and agent developers.

Hardware integration featuring a screen, joystick, and mic for direct interaction with AI agents.

New agent-first retrieval system claiming 4x faster queries and 90% fewer tokens.
Box is utilizing Nvidia models to create a rich context layer for autonomous enterprise agents.
Launch of a quantum error-correction platform integrating HPC and quantum computing.