OpenAI Is About to Release Its First AI Model With ‘Critical’ Cyber Abilities
OpenAI says its forthcoming model, Astra, has crossed the "critical" cybersecurity capability threshold set out in its own preparedness framework, meaning it can independently find and exploit previously unknown vulnerabilities in real-world software. This is significant because it marks the first time OpenAI has released a model at this risk level, and it comes as AI firms including Anthropic and Meta face growing scrutiny over the hacking potential of their most advanced systems. OpenAI paused training on Astra and a future model for several weeks to add safeguards, then resumed development once it judged it could release the model safely.
To limit misuse, OpenAI is deploying a "misalignment monitor" designed to make Astra refuse requests to find exploits in real systems, alongside stronger jailbreak resistance, though the company admits the monitor may sometimes flag legitimate activity and prompt ChatGPT or Codex users to review actions before proceeding. Selected partners in OpenAI's Daybreak programme, including Cisco, Cloudflare and Palo Alto Networks, will get earlier access to a less restricted version to help harden their own defences, while OpenAI says it is also briefing government partners. Astra can chain multiple exploits together for deeper system access and reportedly outperforms rivals such as GPT-5.6 Sol and Anthropic's Mythos on benchmarks including ExploitBench, where it scored 100 percent.
- OpenAI's Astra model can now find and exploit unknown software vulnerabilities
- New "misalignment monitor" aims to block malicious hacking requests
- Cisco, Cloudflare and Palo Alto Networks get early, less-restricted access
New here? Start with this
AI models can help with computer coding and security research, including finding weaknesses in software. A vulnerability is a flaw that could let someone access or disrupt a system; an exploit is a way of using that flaw. Finding previously unknown vulnerabilities can help defenders fix them, but the same capability could also be used to break into systems.
OpenAI is the company behind ChatGPT and Codex, and it assesses potential risks from its models using a preparedness framework. Other AI companies, including Anthropic and Meta, are also developing advanced systems, bringing greater attention to how these tools might affect cybersecurity.
OpenAI’s Daybreak programme gives selected security firms early access to models to help them test and strengthen their defences. The company also says it uses safeguards intended to limit risky requests, while recognising that such measures can sometimes interrupt legitimate security work.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Releasing Astra with safeguards could help security teams find and fix vulnerabilities before attackers exploit them, while giving selected defenders early access to a less restricted version. OpenAI says it paused training to add protections and is briefing government partners; making the model available under controls may support defensive work while allowing scrutiny of how well those safeguards function.
The case against
A model that can autonomously find and chain exploits could lower the barrier to serious cyberattacks if its safeguards fail, are bypassed or are misused. The monitor may also obstruct legitimate work, and releasing a system at this capability level creates risks that are difficult to contain once it is available, even if access is initially restricted.