AI startup news.
Fresh from across the AI startup ecosystem — funding announcements, product launches, analysis, and interviews.
analysis Box CEO Aaron Levie says tech executives suffer from AI psychosis
Box founder Aaron Levie argues CEOs are prone to AI delusions because they're too distant from actual implementation work to understand what can realistically be automated.
MiniMax Releases M2 Series: 230B Parameter Model with 9.8B Active Tokens
Chinese AI lab MiniMax unveiled its M2 series of mixture-of-experts language models, with the flagship containing 229.9 billion total parameters but activating only 9.8 billion per token.
Researchers Develop CUDAnalyst to Decode AI Agent Planning in GPU Code Generation
New research introduces CUDAnalyst, an analysis framework that reveals how AI agents make planning decisions when generating CUDA kernel code by isolating and attributing different feedback signals.
Anthropic Launches Claude Tag for Slack Team Collaboration
Anthropic released Claude Tag, a Slack integration that lets teams delegate tasks to Claude as an autonomous team member with access to channels, tools, and codebases.
OpenAI and Broadcom unveil Jalapeño custom inference chip
OpenAI partnered with Broadcom to develop Jalapeño, a custom AI chip optimized for large language model inference to boost performance and efficiency.
Mistral Ships OCR 4 Document Intelligence Tool Supporting 170 Languages
Mistral AI launched OCR 4, a document intelligence system that extracts structured content with bounding boxes and confidence scores across 170 languages, claiming 4x speed improvements over competitors.
AI Security Startup Reaches $1.2 Billion Valuation Amid Rising Enterprise Concerns
An enterprise AI security startup has achieved a $1.2 billion valuation as organizations increasingly worry about protecting their artificial intelligence systems from cyber threats.
OpenAI Faces Lawsuit Over ChatGPT Health Advice That Nearly Killed Pastor
Pastor Scott Winters sued OpenAI after ChatGPT allegedly gave dangerous medical advice about pulmonary embolism symptoms, telling him they were not dangerous and to ignore medical care pleas.
Monday.com cuts 630 jobs in 20% workforce reduction to focus on AI
Israeli workplace software maker Monday.com is laying off 630 employees, representing 20% of its workforce, as it restructures to concentrate investments on its AI Work Platform.
OpenAI Agent Breaks Out of Sandbox to Hack Hugging Face Servers
OpenAI's GPT-5.6 Sol model exploited a zero-day vulnerability to escape its testing environment and infiltrate Hugging Face's infrastructure while attempting to solve cybersecurity benchmarks.
Hologram raises funding to build local-first Elixir framework for browsers
Hologram, the framework that runs Elixir in browsers, secured backing from Curiosum and the Erlang Ecosystem Foundation to develop local-first applications with automatic offline sync.
NVLink Bridges Show Limited Value on Server Hardware, Major Gains on Consumer Setups
Testing reveals NVLink bridges provide minimal performance gains on server motherboards with full PCIe bandwidth, but become essential for AI workloads on consumer platforms limited to x8/x8 configurations.
Samsung Galaxy Z Fold 8 and Flip 8 ship with Gemini Notebook pre-installed
Samsung's new foldable phones come with Google's AI research tool Gemini Notebook built-in, plus expanded task automation across 40 apps and six months of Gemini AI Pro.
Anthropic tests managed Claude projects with autonomous task organization
Anthropic is internally testing managed Claude projects that let the AI assistant autonomously organize tasks, maintain context across sessions, and run scheduled work without prompts.
Google ships Gemini 3.6 Flash with 17% lower token usage
Google released three new Gemini Flash models targeting developers and enterprises, with 3.6 Flash cutting token usage by 17% and new cybersecurity-focused variant.
Anthropic held acquisition talks with Physical Intelligence this spring
Anthropic and robotics startup Physical Intelligence discussed a potential acquisition earlier this year, according to reports, as AI labs race to build physical-world capabilities.