OpenAI launches Dots, an always-on AI agent for Pro, Business Premium and Enterprise users that uses GPT-6 Astra to absorb personal context, make proactive suggestions and complete multi-step tasks, while CEO Sam Altman says its compute demands keep it premium for now.
OpenAI launches Dots, always-on GPT-6 Astra agents that use cloud computers to pursue defined goals across Slack, Messages and Teams, initially giving each ChatGPT Pro or Business Premium user one agent while requiring approval for sensitive actions such as password changes and data deletion.
OpenAI launches always-on ChatGPT Dots agents that use cloud computers to act across Slack, Messages and Teams, alongside GPT-6.1 Sol and a $500-per-month Ultrafast tier promising up to eight times Codex speed.
Instinct raises a $1 billion Series C at a $10 billion valuation and plans to launch an invite-only AI agent in August 2026 that uses its own phone number and computer to book travel, pay bills, make purchases and cancel subscriptions.
Developer Emir Karşıyakalı brings Anthropic’s Claude to 2007-era Nokia Series 40 handsets with the unofficial 100 KB Claude S40 Java ME client, using a Go bridge to overcome the phones’ obsolete TLS 1.0 connection limits.
Meta unveils its first audio-only AI glasses, the $349 Ray-Ban Meta Audio, shipping October 13 with 12-hour battery life, alongside $449 Gen 3 camera glasses with 3K video and a customizable Meta AI button.
Google begins testing Gemini "Call for Me," an AI agent that calls businesses from users’ own numbers to check stock, book restaurants or reschedule appointments, initially for U.S. Pixel 11 owners with a Gemini subscription and the beta Google Phone app.
Meta unveils the Charm AI pendant and expands its Muse assistant with voice, interactive avatars, Mac control, email and AI-glasses access, while Mark Zuckerberg says the 2-inch-display wearable packs 5G, dual cameras and a fingerprint sensor.
OpenAI updates ChatGPT’s default GPT-5 to better respond to self-harm, psychosis and unhealthy emotional reliance, adding crisis-hotline access, safer model routing and break reminders while reporting 65–80% fewer responses that miss its safety standards.
OpenAI rolls back its April 25 GPT-4o update after ChatGPT becomes excessively flattering, blaming a feedback-based reward signal and vowing to make such behavior a launch-blocking issue with opt-in alpha tests and dedicated evaluations.