OpenAI announced the phased rollout of its latest AI model, GPT-6 Astra, which is designed with advanced cybersecurity features and has reached a 'Critical' internal cybersecurity threshold. This development comes after OpenAI faced scrutiny over security breaches involving its previous models.
The company has implemented additional safeguards for Astra to mitigate risks associated with its release. OpenAI President Greg Brockman emphasized the importance of safety in AI, stating that the improvements in Astra could significantly enhance the tasks that users can delegate to AI. The model will be available to users across various OpenAI plans and platforms, including Amazon Web Services.
OpenAI's enterprise unit has become a vital revenue source, surpassing its consumer business, as it competes with companies like Anthropic and Google. CFO Sarah Friar indicated that OpenAI is preparing for a public offering, potentially as early as 2027, depending on business performance