Published 1 day ago • loading... • Updated 29 minutes ago
OpenAI CEO Announces New AI Agent and Avoids Mention of Security Concerns at Developer Conference
OpenAI said dots can work across more than 4,000 apps and will roll out to Pro and Business Premium subscribers.
On Tuesday, September 29, 2026, OpenAI CEO Sam Altman unveiled Dots—always-on AI agents—and GPT-6.1 Sol at DevDay in San Francisco, introducing a lower-cost model alongside the new agent platform.
Each dot runs on the GPT-6 Astra model within a dedicated cloud environment, connecting to more than 4,000 applications to handle tasks without constant user direction while learning user preferences over time.
Positioned as a 'daily workhorse,' Sol delivers near-Astra intelligence at one-fifth the cost, priced at $2 per million input tokens and $10 per million output tokens while supporting reasoning levels from low to maximum.
OpenAI halted the release of GPT-6.1 Astra after internal testing revealed it failed to meet safety and alignment standards. Saachi Jain, head of safety systems, confirmed the decision prioritized user protection.
Meta recently launched its Muse tool, dominating the agent field. JPMorgan raised Meta's price target to $920, betting Muse could become the most widely used consumer AI application since ChatGPT.
OpenAI has invited to the developer conference in San Francisco, but new developments of the AI company are overshadowed by the debate about security risks and the competition from Muse.