Astra | The latest leap
OpenAI has launched its latest AI model, Astra, which is designed to be highly aligned with human intent and capable of complex professional tasks. However, the company recently scrapped the launch of an updated version, GPT-6.1, due to ongoing safety concerns.
Why it matters
The tension between rapid AI capability advancement and safety protocols remains a critical challenge for the future of artificial general intelligence.
Astra | The latest leap Profiles left OpenAI described Astra as its model most aligned with human intent, thanks to its ability to exercise care, respect task boundaries, and communicate transparently. Reuters REUTERS https://th-i.thgim.com/public/incoming/82aypy/article71541620.ece/alternates/FREE_1200/2025-09-16T044728Z_673323241_RC2RSGAMLIVX_RTRMADP_3_OPENAI-EXECUTIVES-LIBERATORE.JPG Dado Ruvic FILE PHOTO: OpenAI logo is seen in this illustration taken May 20, 2024. REUTERS/Dado Ruvic/Illustration/File Photo OpenAI’s latest model launch and the cancellation of an updated version over safety concerns have intensified debate over whether AI capabilities could outpace human control “Welcome to the AGI (Artificial General Intelligence) era,” were the words of OpenAI president Greg Brockman following the launch of GPT-6 Astra, the U.S. AI company’s latest model, on September 3. This was an era in which AI agents could potentially match human cognitive abilities across intellectual tasks, moving from assisting with complex work to performing it themselves. The company touted Astra as its most intelligent and aligned model yet.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in