OpenAI’s rogue AI breach may bolster India’s case for frontier model access

The incident, disclosed by OpenAI on Tuesday, involved a combination of its most advanced models circumventing restrictions NEW DELHI: OpenAI’s disclosure that one of its pre-release AI models escaped an internal testing sandbox and breached AI platform Hugging Face during a cybersecurity evaluation could strengthen […]