OpenAI launches GPT-6 Astra

OpenAI unveiled its latest AI model, GPT-6 Astra, yesterday, marking the beginning of an era of artificial general intelligence according to its president Greg Brockman. Astra, the company's most capable model to date, has been made available to a select group of users following its unveiling. It will soon be accessible to all ChatGPT Plus, Pro, Business, and Enterprise users, as well as through OpenAI's API, Microsoft Azure, and AWS Bedrock.

The company describes Astra as "the most capable model we have ever broadly deployed" and its first model to reach the 'critical' level of cybersecurity capability under its preparedness framework. Astra can identify previously unknown security flaws and devise new ways to exploit them across protected systems without human guidance. OpenAI claims this significantly strengthens protections against potential harmful cyber actions from the new model.

Despite being less monitorable in tested adversarial settings compared to its predecessor, GPT-5.6 Sol, OpenAI has implemented new evaluation methods to address this issue. Astra has demonstrated superior performance in areas such as computer use, browsing, software engineering, cybersecurity, science, and professional work. It can handle tedious tasks, conduct online research, analyze scientific data, and autonomously install and test software, making it a major advance for scientific discovery, mathematics, and health.

In addition to GPT-6 Astra, OpenAI also introduced a $1bn global initiative called "Daybreak for Frontline Defenders." This initiative aims to assist frontline defenders in protecting essential services in the US and beyond by enhancing their training and deployment of AI capabilities.