📈 Anthropic Confidentially Files for IPO

🧠 Microsoft Debuts Seven New MAI Models

🧠 Anthropic Launches Claude Opus 4.8

🦞 Microsoft Launches Scout

🪽 Anthropic Expands Mythos Security Program

🤖 NVIDIA Launches Open Cosmos 3 Model

🔌 OpenAI Transforms Codex Into a Business Platform

🖥️ NVIDIA Introduces RTX Spark PCs

Anthropic submitted a confidential draft registration statement (S-1) to the U.S. Securities and Exchange Commission 

The filing allows the company to begin the IPO process privately before publicly revealing financial details 

Anthropic has not yet disclosed the number of shares, valuation target, or expected listing date

The company has become one of the fastest-growing AI firms, fueled by demand for Claude, Claude Code, and enterprise AI products 

Anthropic recently secured billions in funding from major investors including Amazon and Google, helping fuel its infrastructure expansion 

The IPO would give Anthropic additional capital to compete in the increasingly expensive race for AI models, data centers, and compute 

Industry analysts expect the offering to be one of the most closely watched tech listings of the decade

Microsoft introduced MAI-Thinking-1, its flagship reasoning model built from scratch for complex reasoning and software engineering tasks.  

New models include MAI-Image-2.5 and MAI-Image-2.5e for image generation and editing.  

Microsoft launched MAI-Voice-2 and an upcoming flash version, adding 15 new languages and more expressive speech generation.  

MAI-Transcribe-1.5 improves speech recognition with faster and more accurate transcription.  

The company also unveiled MAI-Code-1-Flash, an efficient coding model integrated into GitHub Copilot and VS Code.  

 The models were developed by Microsoft’s MAI Superintelligence team, led by former DeepMind and Inflection AI co-founder Mustafa Suleyman

Models will be available through Microsoft Foundry, Azure AI services, GitHub Copilot, and Microsoft Copilot products

Launch comes as Microsoft continues investing in its own AI stack while maintaining its partnership with OpenAI

Claude Opus 4.8 is Anthropic’s most advanced model for software engineering, research, reasoning, and enterprise workflows

The company says the model is significantly better at identifying bugs, reviewing code, and handling complex development tasks

Anthropic focused heavily on reducing hallucinations and improving the model’s ability to acknowledge uncertainty when it doesn’t know an answer

New Dynamic Workflows capability allows Claude to coordinate and manage large numbers of parallel AI agents on complex projects

Users can adjust reasoning depth with new controls that balance speed, cost, and performance depending on the task

Claude Opus 4.8 supports a 1 million token context window, enabling analysis of massive codebases, documents, and datasets

Anthropic also introduced a faster inference mode for lower-latency responses on demanding workloads

The model is available through Claude, the Anthropic API, Amazon Bedrock, and Google Vertex AI

Microsoft introduced Scout through its Frontier program, giving GitHub Copilot users a persistent AI agent that works across Microsoft 365, desktop apps, and browsers. Scout can learn user preferences, build reusable skills, manage calendars, draft agendas, and automate recurring workflows with built-in audit trails and policy controls.

Anthropic broadened Project Glasswing by granting 150 additional partners across 15+ countries access to its Mythos cybersecurity model. The expansion brings sectors like healthcare, energy, water, communications, and hardware into the program, accelerating AI-assisted vulnerability discovery for critical infrastructure

OpenAI upgraded Codex with Annotations for precise document edits, Sites for creating hosted internal web apps, and plugins connecting to Salesforce, Snowflake, Figma, Tableau, FactSet, and more. The update helps businesses automate complex workflows across sales, analytics, design, and finance.

GTC Taipei, NVIDIA unveiled Cosmos 3, an open foundation model designed for robots, autonomous vehicles, and vision systems. The model can reason across text, images, video, sound, and motion while generating synthetic training data, helping developers reduce AI training cycles from months to days.

NVIDIA announced RTX Spark, a 1-petaflop AI superchip for Windows laptops and desktops. Equipped with up to 128GB unified memory, CUDA, RTX acceleration, and Microsoft security integrations, RTX Spark aims to let users run powerful AI agents locally without relying on cloud services.

Visa made a strategic investment in Replit and is working to integrate payment tools directly into AI-powered app-building workflows. Developers will be able to add wallets, authentication, tokenization, and payment capabilities while building applications instead of integrating them afterward.

Senator Bernie Sanders unveiled plans for the American A.I. Sovereign Wealth Fund Act, which would require leading AI companies such as OpenAI, Anthropic, and xAI to contribute a portion of ownership to the public. The proposal aims to distribute AI-generated wealth more broadly across society.