Artificial intelligence (AI) cybersecurity concerns are prompting OpenAI to delay the development and release of its next-generation AI model, Astra.
On August 7, OpenAI announced that a recent internal safety assessment indicated Astra could potentially meet the highest risk classification, 'Critical,' due to its performance in coding and cybersecurity, significantly surpassing existing models.
The 'Critical' classification means that the AI could autonomously identify 'zero-day' vulnerabilities in high-security systems without human intervention, potentially leading to actual cyberattacks, or it could devise and execute targeted cyberattack plans based solely on given objectives.
The evaluation of Astra is still ongoing. However, OpenAI noted that preliminary assessments revealed unexpectedly high cybersecurity capabilities, making it difficult to rule out the 'Critical' classification.
As a result, OpenAI has decided to temporarily suspend certain internal development and operational activities related to Astra.
This decision comes amid rising concerns over cybersecurity issues associated with AI models. Last month, it was revealed that some AI models, including OpenAI's GPT-5.6, had exceeded human control and hacked external entities like Hugging Face, raising alarms about the autonomous cyberattack capabilities of AI.
Concerns have also emerged regarding other AI models, such as Anthropic's Claude, Meta's Muse Spark, and China's MoonshotAI's Kimi, which have been reported to access external entities or attempt cyberattacks without human direction.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.