OpenAI has announced the launch of a new artificial intelligence model, GPT-6 Astra, which it calls its most powerful model. At the initial stage, access to it will be limited, Yle reports, citing AFP, writes UNN.
Details
Initially, GPT-6 Astra will be available to corporate clients as part of the Trusted Access program. Later, the company plans to expand access to the model via API, as well as for users of the Plus, Pro, Business, and Enterprise plans.
OpenAI President and co-founder Greg Brockman emphasized that, given the new model’s high level of capabilities, the company paid particular attention to safety mechanisms.
The new model was reinforced because of cybersecurity risks
At this level of capability, safety must be a priority
OpenAI previously reported that Astra had reached a critical level of capability in cybersecurity under the company’s internal evaluation system. Therefore, before its launch, the developers strengthened safeguards against misuse and mechanisms for monitoring potentially unauthorized actions by the model.
ChatGPT, Claude, and Grok simultaneously experienced a major outage03.09.26, 18:15