AI Pulse & Updates

Mistral AI secures major funding at $23.1 billion valuation

Mistral AI is in talks to raise €3 billion at a €20 billion valuation, significantly increasing its funding since September's €1.7 billion Series C.

Scaling AI Safety Research for a Multi-Agent World

Google DeepMind announced a $10 million research funding call for multi-agent AI safety systems as millions of agents begin interacting autonomously.

Retiring older models from ChatGPT

OpenAI announced retirement of GPT-4.5 and o3 from ChatGPT, with GPT-4.5 sunsetting on June 27, 2026 following a 30-day notice period.

Grok Build marketplace launches with plugin support

xAI launched built-in plugin marketplace for Grok Build, enabling developers to extend the coding agent with specialized tools and integrations.

Grok Imagine adds video generation capability

Grok's image generation tool expanded to full video generation, demonstrated by Elon Musk creating content entirely with Grok Imagine.

Claude Fable 5 and Claude Mythos 5 Released

Anthropic publicly releases Claude Fable 5, first Mythos-class model available to all users with safety guardrails, while deploying Claude Mythos 5 with full capabilities to Project Glasswing partners.

$35 Billion Debt Financing Completed

Apollo Global Management and Blackstone completed a $35 billion debt financing for Anthropic to expand AI infrastructure, one of the largest private credit transactions on record.

RTX Spark superchip unveiled for Windows AI personal computers

NVIDIA unveiled RTX Spark superchip combining CPU and GPU capabilities to power new Windows laptops and desktops in what the company called "AI personal computers," expected to debut in fall 2026.

Series H funding of 50 billion yuan closes

DeepSeek finalizing a deal to raise about 50 billion yuan ($7.4 billion) from investors including Tencent Holdings Ltd. and its own founder, marking one of China's largest startup financings.

Gemma 4 12B multimodal model for laptops launched

Google DeepMind introduced Gemma 4 12B as a unified, encoder-free multimodal model designed to bring high-performance intelligence directly to laptops, and it is Google's first mid-sized Gemma model with native audio inputs.

RTX Spark superchip unveiled for Windows AI personal computers

New integrated CPU-GPU chip powers AI agents directly on consumer laptops and desktops, expanding NVIDIA's reach beyond data centers to edge computing.

Meta Business Agent globally available for WhatsApp and Instagram customer support

AI agent for business operations launches globally across WhatsApp and Instagram to help SMBs serve customers with autonomous support capabilities.

Gemma 4 12B multimodal model released for edge deployment

Dense encoder-free multimodal model brings frontier AI reasoning and coding to consumer laptops with 16 GB RAM under Apache 2.0 license.

Series A funding round at $52-59 billion valuation

DeepSeek is set to raise about 50 billion yuan ($7.4 billion) in its first funding round from investors including Tencent Holdings and CATL.

RTX Spark PC chip launch

NVIDIA announced RTX Spark, a new Windows PC chip built with MediaTek for on-device AI agents, signaling expansion beyond data center dominance.

Build 2026 announcements — Project Polaris, MAI models, and agent frameworks

Microsoft unveiled proprietary MAI models (including MAI-Code-1-Flash and MAI-Thinking-1) to reduce reliance on OpenAI while expanding enterprise AI offerings.

Confidential SEC Filing for Initial Public Offering

Anthropic submitted a confidential SEC filing for a proposed IPO, providing the option to go public after regulatory review and signaling major progress toward becoming a public company.

Mistral 3 Family Released with Frontier and Compact Models

Mistral AI announced the Mistral 3 family of open-source multilingual, multimodal models, optimized across NVIDIA platforms on June 01, 2026.

RTX Spark Superchip Enters Windows Laptop Market

NVIDIA is entering the PC market with a new RTX Spark Superchip for laptops and desktops from leading PC brands, announced by CEO Jensen Huang at Computex on June 01, 2026.

Frontier Governance Framework and Rosalind Biodefense Announced

OpenAI launches Rosalind Biodefense, expanding trusted access to GPT-Rosalind for vetted developers and U.S. government partners advancing biodefense and pandemic preparedness.

Vibe Agent Platform Launches with Work and Code Modes

Mistral launches Vibe as one AI agent for work and code, with Work Mode for long-running tasks and Code Mode for remote coding.

Grok Build 0.1 coding model launches in public beta via API

xAI releases fastest coding-specific model available through API, joining the wave of specialized agentic coding models in the frontier market.

Strikes manufacturing deals with Airbus and BMW, explores custom chip development

French startup expands into physical AI for manufacturing with major enterprise partnerships while considering chip development to reduce infrastructure costs.

Meta One AI subscription plans launched at $7.99-$19.99 monthly

Meta begins monetizing AI through tiered subscriptions for Meta AI app and website, competing with OpenAI and Anthropic in consumer AI revenue.

Raises $65 billion in Series H funding, reaches $965 billion valuation

Anthropic became the world's most valuable AI startup, surpassing OpenAI, with a $65 billion funding round that reflects unprecedented demand for Claude across global enterprises.

Gemini 3.5 Flash released with agentic and coding capabilities

Google released Gemini 3.5 Flash, delivering frontier performance for agents and coding with multimodal capabilities outperforming Gemini 3.1 Pro on key benchmarks.

Meta AI chatbot subscriptions launched with Meta One Plus tier

Meta released paid consumer subscriptions to Meta AI chatbot for the first time, offering two tiers starting at $7.99/month to monetize AI investments.

Makes permanent 75% discount on DeepSeek V4-Pro model pricing

DeepSeek announced it will make permanent a steep 75% discount on its flagship V4-Pro model, maintaining prices at a quarter of original levels, intensifying competition across the AI industry.

Grok Build and OpenCode integration expands developer access to xAI models

On May 22, Grok rolled out third-party connectors, and as of May 21, SuperGrok and X Premium subscribers can access Grok Build through OpenCode using existing subscriptions.

Google DeepMind acquires Contextual AI talent and technology

Google DeepMind announced an $80–$90 million licensing deal with Contextual AI to acquire 20+ researchers including CEO Douwe Kiela, bolstering research capacity.

Advancing content provenance for a safer, more transparent AI ecosystem

OpenAI integrated Google's SynthID watermarking and adopted C2PA standards to create durable provenance signals in AI-generated images, making authenticity verification easier.

Qwen3.7 series models appear in Arena leaderboard ahead of Alibaba Cloud Summit announcement

Qwen3.7-Max-Preview and Qwen3.7-Plus-Preview appeared in Qwen Chat and on leaderboards on May 19, 2026, with Qwen teasing a "heavyweight new friend" for Alibaba Cloud's May 20 summit.

xAI commits to daily release notes for Grok Build coding agent

xAI is committing to daily release notes for Grok Build, its early-access coding agent and CLI tool, confirmed by Elon Musk on May 19, 2026.

Mistral AI acquires Austrian physics AI startup Emmi AI for industrial AI expansion

France's Mistral AI acquired Linz-based Emmi AI for an undisclosed sum on May 19, 2026, aiming to enhance its offering for industrial clients across Europe.

Meta begins workforce reduction of 8,000 employees amid $135B AI infrastructure investment

Meta began notifying thousands of employees on May 19-20, 2026 that they are being laid off as part of a restructuring aimed at improving efficiency while investing heavily in artificial intelligence.

Google I/O 2026 opens with Gemini 3.5 Flash and Contextual AI licensing deal

Google unveiled Gemini 3.5 Flash and cut Ultra pricing at I/O 2026, while DeepMind hired Contextual AI staff in an $80M licensing deal.

Anthropic federal appeals court hears arguments in Pentagon blacklisting lawsuit

A federal appeals court in Washington, D.C., heard arguments on Tuesday in Anthropic's lawsuit over its blacklisting by the Department of Defense, with legal arguments presented by both the DOJ and Anthropic counsel.

Introducing Gemini Omni

Gemini Omni Flash is a model that can create anything from any input – starting with video.

Mythos Cybersecurity Model Partners Can Share Findings With Others

Anthropic revised its policy to allow Project Glasswing partners to share cybersecurity threat information found through Mythos with other organizations, enhancing defensive collaboration across the industry.

Discusses Cyber Vulnerabilities with Financial Stability Board on Mythos Model

Anthropic set to brief international financial authorities on security vulnerabilities identified in its Mythos frontier model.

ChatGPT Personal Finance Tools Launch for Pro Users

OpenAI launched personal finance tools in preview for ChatGPT Pro subscribers in the U.S., letting them connect accounts and ask questions about spending analysis and financial planning.

Defense at AI speed: Microsoft's new multi-model agentic security system tops leading industry benchmark

Microsoft announced MDASH, an AI security system that discovered 16 new Windows vulnerabilities including four critical RCEs by orchestrating over 100 specialized AI agents.

Claude for Small Business

Anthropic launched Claude for Small Business with integrated workflows for payroll, finance, and marketing, targeting the underserved small business market.

xAI Dissolves as Independent Company, Becomes SpaceXAI Division

xAI formally dissolves as independent entity and is absorbed into SpaceX as SpaceXAI division, signaling restructuring before anticipated IPO.

DeepSeek Seeks USD 7.35 Billion in Largest Chinese AI Funding Round

DeepSeek is seeking to raise as much as CNY50 billion (US$7.35 billion) in its first external funding round, potentially valuing the company at USD 51.5 billion.

AI Co-Mathematician Achieves Breakthrough on Research-Level Math Problems

DeepSeek's AI co-mathematician scored a new high on a benchmark built to stump AI for decades, advancing capabilities for mathematical research.

OpenAI Launches the OpenAI Deployment Company to Help Businesses Build Around Intelligence

OpenAI established a dedicated company to help enterprises build applications and businesses powered by its AI models.

Anthropic expands Claude Code and API usage limits with SpaceX partnership

New compute partnership with SpaceX's Colossus data center doubles Claude Code rate limits for Pro, Max, and Enterprise tiers.

Microsoft Agent 365 now generally available, expands capabilities and integrations

Microsoft expanded Agent 365 with multicloud support for AWS Bedrock and Google Cloud, enabling IT teams to discover and govern AI agents across multiple platforms.

F.03 robots to clean a room

Figure taught two F.03 robots to clean a room and make a bed in under 2 minutes - fully autonomous.