Anthropic introduced Claude Sonnet 5.5, over 30% faster and up to 30% cheaper than Sonnet 5 for most work
Codex drove 64% of OpenAI’s enterprise output tokens, up from near zero a year ago
Perplexity gave 9 AI models root access in a sandbox test, and none breached the VM boundary across 108 runs and joined Nvidia and 100+ industry partners to build infrastructure that contains rogue AI agents.
Nvidia and 100+ partners launched an open safety platform to monitor and enforce boundaries on AI agents, bringing together OpenShell and Sentry
Meta launched Enterprise Platform with its AI stack (Muse agent, Meta Business Agent, Muse API, Muse Code), led by former MongoDB CEO Chirantan Desai
ElevenLabs launches Eleven v4 and Eleven v4 Turbo, with v4 taking first place on the Artificial Analysis Provider Voice TTS Arena
TinyFish launches two free builder programs, one for students and one for ambassadors, sharing a monthly bounty board where multiple users can get paid for the same bounty
Manus launches Manus 2.0 with Cloud Computers, event-triggered Automations, a Manus Studio desktop workspace, and Cue, a standalone app for personal agents with their own email, phone number, and wallet
Base44 launches Base Code, a cloud code editor that opens any GitHub repo in the browser, works with multiple AI models side by side, and turns queries into pull requests
Grok Bot introduced Team Bots as shared AI teammates that learn as your team works with them, usable in Slack or Grok Bot
World Labs is joining AMD to scale its spatial and physical intelligence work closer to the hardware
Samsung commits $1B to AI infrastructure company Helix, on top of the $10B secured when a KKR-led consortium including Nvidia set it up in June
Rep. Ro Khanna will introduce the Human Control Over AI Act, with strict liability and a ban on recursive self-improving AI until government safeguards exist