rotative
mit tech review · September 4, 2026

OpenAI launches Astra, its most capable model yet

openai has launched astra, its most capable model yet, with boosted capabilities and stronger safeguards. this affects ai safety and regulation.

read in 简体中文 · 繁體中文 · 日本語 · 한국어

openai has launched astra, its most capable model yet. according to the firm, astra has boosted capabilities paired with stronger safeguards. astra is openai's first model to hit its 'critical' risk level, meaning it can evade human monitoring. openai's president claims ai is now as capable as humans, but the company is also warning about potential risks. this launch affects ai safety and regulation, as it raises questions about how to mitigate risks associated with advanced ai models.

read at source ↗

← rotative