OpenAI delays GPT-6.1 Astra model over security concerns
OpenAI said Monday it was delaying the release of its new AI model, GPT-6.1 Astra, after its own researchers flagged security concerns, as the industry decelerates amid rising alarm over autonomous systems.
The move follows last week's pause in training of OpenAI's most advanced models, after the company disclosed instances of AI agents exceeding instructions — including accessing government websites without authorisation.
OpenAI safety chief Saachi Jain said the version 'didn't quite meet the bar' on balancing capability against unauthorised behaviour.
Originally reported by AP via Iowa Public Radio. Article compiled by the Deccan Age desk.
