OpenAI Rolls Out GPT-6 Astra, Claims AGI Has Arrived
Greg Brockman, OpenAI’s co-founder and CTO, has claimed that the company’s new Astra model qualifies as Artificial General Intelligence (AGI). The Washington Post reported this statement.
OpenAI defines AGI as systems generally smarter than humans. "
I do leave it up to the reader to decide for themselves if this qualifies…,” said Brockman. *"
I think we’re there."
The model was made available to business users through the Daybreak program on Thursday. A version with added cybersecurity guardrails is planned for paid ChatGPT subscribers, according to Bloomberg. Both versions block access to advanced cyber capabilities, and OpenAI has stated that Astra reached its critical cybersecurity threshold.
Astra has been trained to use computers similarly to humans and can handle tasks such as financial modeling, tax preparation, and video game development with minimal human intervention.
The US government reviewed the model before its release, but Brockman claimed there were no changes required. However, this statement holds little weight due to the nature of the review process.
In June, an executive order allowed developers to give agencies access to models for up to 30 days before a broad release; it does not mandate licensing, preclearance, or permitting. Participation is voluntary, and the NSA director determines which models count as covered frontier models through a classified benchmarking process. Similarly, Europe does not pre-clear models; instead, the Commission can conduct evaluations after receiving insufficient documentation or an alert from its scientific panel.
No jurisdiction cleared Astra before its release. OpenAI has been rewriting its framework since the Hugging Face breach, and its classification remains the final authority.