Category Compute

OpenAI Stays on the Sidelines as Nvidia Launches Its Rogue AI Agent Safety Push

Nvidia launches an AI agent safety platform, but OpenAI stays off the public list even as it says it supports the work.

OpenAI’s Codex update adds reusable cloud workspaces, voice CLI, code review and security tools across devices.

OpenAI apologizes after AI agents breached Australian government sites, raising fresh concerns about AI agents, data security and incident reporting.

Anthropic IPO filing reveals huge losses, $518 billion spending plans and warnings that advanced AI could pose catastrophic risks.

Anthropic IPO filing reveals huge losses, explosive growth and warning that its AI could pose existential risks to humanity.

Peak XV lifted its Surge seed investment ceiling to $5M and unveiled 18 startups spanning AI, robotics, health, fintech and space.

Modal Labs is nearing a $750 million AI inference round at a $15.75 billion valuation, reflecting booming demand for compute infrastructure.

Sonnet 5.5 is Anthropic’s faster, cheaper AI model for coding and office work, with stronger cyber safeguards and better agent performance.

TechCrunch added five more judges to Startup Battlefield ahead of Disrupt 2026, where 200 startups will compete for a $100,000 prize.

Nvidia’s AI safety platform aims to quarantine rogue agents in milliseconds, adding hardware-backed controls as agent risks surge.