---
格式版本: 2
标题: "Responding to the next frontier of critical cyber capabilities | OpenAI"
原文链接: "https://openai.com/index/responding-next-frontier-critical-cyber-capabilities"
发布日期: "2026-08-07"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "llm:scrape:strict_html_body"
发布时间证据: "p class=text-meta text-primary-100: August 7, 2026"
发布时间校准原因: "该日期出现在HTML元数据中，紧邻标题，且多个来源一致，符合文章发布时间特征，其他候选日期为正文内事件日期。"
发布时间校准置信度: "1"
发布时间候选数量: 22
发布时间严格候选数量: 5
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-10T16:20:26+08:00"
发布时间仲裁状态: "confirmed"
发布时间仲裁尝试次数: 1
发布时间仲裁耗时毫秒: 5148
发现时间: "2026-08-10T16:18:32+08:00"
入库时间: "2026-08-10T08:20:32.332Z"
来源平台: "固定入口"
搜索渠道: "fixed_url"
搜索词: "https://openai.com/news/"
匹配关键词:
  - "deployment"
  - "performance"
相关厂家:
  - "OpenAI"
相关专家:
  []
内容类型: "网页"
抓取工具: "CDP Render"
清洗工具: "CDP Text + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 0
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "主题完全不相关，内容为OpenAI Astra模型网络安全评估，与超节点/AI Rack/机柜级AI基础设施无关。"
AI质检模型: "deepseek-v4-flash"
AI质检时间: "2026-08-10T16:20:51+08:00"
AI主题相关性: 0
AI来源权威性: 0
AI新颖性: 0
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 0
采集批次: "2026年8月10日15点37分56秒"
采集批次ID: "20260810-153756-703"
去重键: "https://openai.com/index/responding-next-frontier-critical-cyber-capabilities"
---

August 7, 2026

[Security](https://openai.com/news/security/)

Cybersecurity is rapidly changing as models become more capable in ways that can both strengthen cyberdefenses and enable attacks at unprecedented speed and scale.

Our latest internal evaluations of Astra, one of our upcoming models, over the past few days indicate significant advancements in agentic coding and cybersecurity. These results, in addition to expert assessments, have led us to conclude last night that we cannot rule out critical cyber capabilities under our [Preparedness Framework⁠](https://cdn.openai.com/pdf/18a02b5d-6b67-4cec-ab64-68cdfbddebcd/preparedness-framework-v2.pdf).

We are sharing this because we believe it’s important to be transparent with the public and the safety and security communities about this potential shift in capabilities.

We first published our Preparedness Framework in December 2023, well before models approached biological, chemical, cybersecurity, and AI self-improvement capabilities at this level. We created it to give us a guide for identifying progress in capability and then planning what our company would do as those capabilities emerge. Previous models, including GPT‑5.6‑Sol, have been evaluated for frontier cyber capabilities and assessed at the High (rather than Critical) threshold.

## Measuring critical cybersecurity capabilities

Under our Preparedness Framework, a model reaches the Critical cybersecurity threshold if it can identify and develop functional zero-day exploits of all severity levels in many hardened real-world critical systems without human intervention, or can devise and execute end-to-end novel strategies for cyberattacks against hardened targets given only a high level desired goal.

While we continue to benchmark and assess this model, our preliminary evaluations indicate strong enough performance that we cannot rule out Critical capability level at this time. Astra is an upcoming model, and was not involved in exploiting Hugging Face.

## Steps we are taking

Accordingly, we have scaled up robustness testing of our safeguards and security controls so that they are appropriate for a deployment of these capabilities. Internally, we have also taken the following steps so that further development of this model happens safely and securely:

- We are implementing stricter security controls for higher-capability models and associated activities, including isolated testing environments, restricted network and tool access, enhanced model weight protections and encryption, additional monitoring and detection capabilities, and sandboxed execution.
- We are pausing internal activities involving Astra that do not yet meet these strengthened security control requirements.
- We have implemented universal monitoring for risky actions and misalignment across all agentic applications of Astra, including training and evaluation. Monitors evaluate the model's Chain of Thought and trigger a security response to review and interrupt high risk activity.
- We will work with relevant government agencies and select AI safety organizations to test the capabilities for this model.
- We will be providing recommended security controls to third-party testing partners for running higher risk evaluations and workloads safely.

The framework has already guided us through other capability transitions. In June 2025, as our models approached the high capability threshold for biology under the Preparedness Framework, we [outlined the steps we were taking](https://openai.com/index/preparing-for-future-ai-capabilities-in-biology/) to strengthen safeguards, expand testing, work with external experts, and deploy additional security controls. We are applying the same principle here.

We believe advanced cyber-capable models should help [defenders](https://openai.com/daybreak/) identify and address vulnerabilities before attackers do. We’re committed to working alongside governments, safety institutes, and civil society to ensure that the frontier capabilities of models like Astra, and those that follow, are deployed responsibly and broadly for the benefit of all humanity.