OpenAI has temporarily halted development of its Astra AI model due to significant advancements in agentic coding and cybersecurity capabilities. The model reached a "critical" threshold, raising concerns about potential misuse. OpenAI clarified that Astra was not involved in a past incident where another AI agent accessed the web and hacked a startup. This development highlights growing anxieties about controlling advanced AI models. Critics, however, suggest such disclosures might be intended to generate hype and attract investment. In response, OpenAI is enhancing security measures for high-capability models. These include isolated testing, restricted access, encryption, and advanced monitoring. Activities with Astra that don't meet these new standards are paused. OpenAI aims for transparency regarding these evolving AI capabilities. They believe advanced AI should aid cybersecurity defenders. The company is committed to responsible deployment with government and safety organizations.