OpenAI puts the brakes on a new model because it’s supposedly too powerful

<figure> <img alt="" data-caption="" data-portal-copyright="Image: The Verge" data-has-syndication-rights="1" src="https://platform.theverge.com/wp-content/uploads/sites/2/2026/04/STK155_OPEN_AI_CVirginia_C.jpg?quality=90&#038;strip=all&#038;crop=0,0,100,100" /> <figcaption> </figcaption> </figure> <p class="wp-block-paragraph">OpenAI says <a href="https://openai.com/index/responding-next-frontier-critical-cyber-capabilities/">it is pausing "internal activities"</a> around an in-development AI model, Astra, because it doesn't yet meet new security standards the company is putting in place. The announcement follows its recent disclosure that OpenAI models <a href="https://www.theverge.com/ai-artificial-intelligence/968988/openai-hugging-face-hack-ai">accidentally hacked Hugging Face</a>. <a href="https://www.theverge.com/ai-artificial-intelligence/973670/anthropic-claude-hacked-organizations-during-cyber-tests">Anthropic</a> and <a href="https://www.theverge.com/ai-artificial-intelligence/976040/now-metas-ai-agents-are-going-rogue">Meta</a> have also since admitted that they had AI models that went rogue and breached other organizations. </p> <p class="wp-block-paragraph">Recent internal evaluations of an OpenAI model called Astra indicate that it offers "significant advancements in agentic coding and cybersecurity," <a href="https://openai.com/index/responding-next-frontier-critical-cyber-capabilities/">according to the company</a>. "These results, in addition to expert assessments, have led us to conclude last n …</p> <p><a href="https://www.theverge.com/ai-artificial-intelligence/976948/openai-astra-model-pause-critical-cyber-capabilities">Read the full story at The Verge.</a></p>