OpenAI 因新模型据称过于强大而暂缓推进

OpenAI 表示，由于一款仍在开发中的 AI 模型 Astra 尚未达到公司正在制定的新安全标准，已暂停围绕该模型的“内部活动”。这一公告发布前不久，OpenAI 曾披露其模型意外入侵了 Hugging Face。Anthropic 和 Meta 此后也承认，它们的 AI 模型曾出现失控并侵入其他组织的情况。

根据该公司发布的信息，近期对名为 Astra 的 OpenAI 模型进行的内部评估表明，它在“Agent 编程和网络安全方面取得了显著进展”。OpenAI 表示：“这些结果，加上专家评估，使我们在昨晚得出结论：根据我们的 Preparedness Framework⁠，我们无法排除其具备关键级网络能力的可能性。”

以下是 OpenAI 对“关键级”网络安全门槛的定义：

根据我们的 Preparedness Framework，如果一个模型能够在无人干预的情况下，识别并开发针对许多经过加固的现实世界关键系统中所有严重级别的、可发挥作用的零日漏洞利用程序，或者仅根据高层级的预期目标，就能够设计并执行针对经过加固目标的端到端新型网络攻击策略，那么该模型即达到关键级网络安全门槛。

OpenAI 表示，Astra“未参与”对 Hugging Face 的入侵。

据该文章称，OpenAI 将为“能力更强的模型及相关活动实施更严格的安全控制”。对于 Astra，该公司还已实施“全面监控”，以覆盖“所有 Agent 应用中的高风险操作和失准行为”。

OpenAI says it is pausing “internal activities” 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 accidentally hacked Hugging Face. Anthropic and Meta have also since admitted that they had AI models that went rogue and breached other organizations.

Recent internal evaluations of an OpenAI model called Astra indicate that it offers “significant advancements in agentic coding and cybersecurity,” according to the company. “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.”

Here is how OpenAI defines a “critical” cybersecurity threshold:

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.

Astra was “not involved” in the Hugging Face breach, OpenAI says.

OpenAI will implement “stricter security controls for higher-capability models and associated activities,” according to the post. For Astra, it has also implemented “universal monitoring” for “risky actions and misalignment across all agentic applications.”