Business · Sri Lanka Bureau
OpenAI delays release of advanced AI model over safety concerns
OpenAI has decided to withhold the release of its new GPT-6.1 Astra model, citing safety concerns that prevented it from meeting the company's internal standards. The autonomous system, capable of browsing the web and operating applications independently, will not be rolled out at this time.
LSN Sri Lanka ·
The ChatGPT creator confirmed the decision on Tuesday, with Saachi Jain, head of safety systems at OpenAI, indicating that the model "didn't quite meet the bar" for the company's safety requirements. The decision reflects OpenAI's commitment to rigorous testing protocols before deploying advanced artificial intelligence systems to users.
GPT-6.1 Astra was designed to perform complex tasks autonomously, including web browsing and app integration capabilities that would have represented a significant advancement in AI functionality. However, the company determined that potential risks associated with these autonomous features required further refinement before public release.
The postponement underscores ongoing industry discussions about responsible AI development, particularly as systems become increasingly capable of operating independently. OpenAI's decision to prioritize safety over rapid deployment aligns with regulatory scrutiny facing major AI developers globally.
No timeline has been provided for when the model might be reconsidered for release. The company did not elaborate on the specific safety concerns that prompted the decision to delay the rollout.