< Back to all clusters
[TECHNOLOGY] · United States, United Kingdom · 100 sources

started · updated

OpenAI launches GPT-6 Astra with autonomous computer use capabilities

OpenAI has released GPT-6 Astra, a model described by President Greg Brockman as a potential milestone for Artificial General Intelligence (AGI). Released on September 3, 2026, the model features a 1.05 million token context window and advanced autonomous capabilities, allowing it to operate computers, navigate browsers, and execute multi-step professional tasks such as software engineering and web development.

Due to its advanced capabilities, Astra is the first OpenAI model to trigger the company's strictest safety protocols. OpenAI executives, including Amelia Glaese, noted that the model can autonomously identify and exploit previously unknown security vulnerabilities in well-protected systems. This high level of capability has led to a cautious deployment, initially targeting specific organizations before expanding to Plus, Pro, Business, and Enterprise users, as well as via API and Amazon Web Services.

In response to the rapid advancement of such models, OpenAI Chief Scientist Jakub Pachocki has called for the industry to adopt mandatory international safety standards and suggested that laboratories may need to voluntarily slow development to manage the risks associated with increasingly autonomous AI agents.

Entities

Amelia Glaese · Anthropic · GPT-6 Astra · Greg Brockman · Jakub Pachocki · Meta · Microsoft · OpenAI · Sam Altman

Claims

What the coverage asserts, and how many sources carry each claim.

Sources

9 days ago
9 days ago
8 days ago
9 days ago
8 days ago
9 days ago
7 days ago
8 days ago