
OpenAI valued at 1.25 trillion dollars as rival Anthropic nears IPO
OpenAI has reached a 1.25 trillion dollar valuation after raising another 11 billion dollars, while its competitor Anthropic moves closer to an IPO.
Category
AI models, automation and artificial intelligence news.
1656 articles

OpenAI has reached a 1.25 trillion dollar valuation after raising another 11 billion dollars, while its competitor Anthropic moves closer to an IPO.

Chinese intelligence reportedly recruited Megaspeed executive Alice Huang to spy on Nvidia CEO Jensen Huang and gather AI chip data.

Tesla unveils its updated AI accelerator roadmap, renaming chips to SI. Sixth-gen SI6 chips will deliver an 80x performance increase.

University of Michigan researchers discovered that AI processing queries in various national languages requires up to 180 times more energy.

Meta introduces Spaces, a new feature for Muse AI agents that allows users to create dedicated project boards, trackers, and automated workspaces.

ChatGPT users are reporting that the service automatically reverts their selected reasoning level from High to Medium, ignoring user settings.

Meta is rolling out spaces in its Muse AI assistant, allowing users to build customized, app-like experiences directly in the app.

XDA tests ChatGPT Dots, comparing it with OpenClaw and Perplexity Computer. Discover how OpenAI's new proactive AI personal assistant works.

The US Department of Justice has launched the Super Intelligence Task Force (DOJ-SITF) to protect AI technologies from malicious actors.

Google is developing a new Gemini feature that will alert parents if minor users discuss self-harm or hurting themselves with the chatbot.

OpenAI has unveiled GPT-6 Cyber, a new security-focused AI model designed to find and fix software vulnerabilities while maintaining strict alignment.

Google launches Gemini Omni 1.2 Flash featuring 15-second native video generation, 30% lower pricing, and multi-turn conversational editing.

Mira Murati's 594B parameter Inkling 2 Small model enters the top three in DesignArena for website generation.

A new Ramp report reveals AI spending flatlined over late summer despite rising token usage, challenging hypergrowth assumptions ahead of Anthropic's IPO.

An independent study by METR reveals that AI integration has nearly doubled the development pace of Anthropic models since October 2025.

An XDA writer successfully used the Codex AI agent to automatically migrate, categorize, and tag 1,000 notes from Apple Notes, Evernote, and Google Keep into Obsidian.

Google DeepMind launched Gemini Omni 1.2 Flash for video generation, dropping the price to $0.07 per second and adding an SLA.

Tencent has released the final version of its flagship Hy4 model, featuring GUI and CLI computer control capabilities along with image analysis.

OpenAI engineer teases Le GPT Fat while dual VMs appear in the API. How 10,000 autonomous agents tackled the Navier-Stokes equations.

DeepSeek V4.1 Flash is now available in Selectel Token Factory. A model with one million context tokens running on local infrastructure.

Physicl has secured $10M in seed funding to create high-fidelity 3D environments, helping robots learn to interact with the physical world more effectively.

Startup Underdog has released Woof, a 31.7MB speech and music model that runs locally inside the browser without requiring internet.

A University of Chicago law professor warned about AI risks after a Claude AI agent mistakenly deleted a full year of lecture notes.

OpenAI is acquiring Pierre Computer, a developer of Git tools and code storage infrastructure. The tech will be integrated into Codex.