内容来源：https://techcrunch.com/2026/06/25/the-white-house-is-asking-openai-to-slow-roll-the-release-of-its-new-model-over-safety-concerns/

内容总结：

美政府施压OpenAI：新一代模型GPT-5.6仅限合作伙伴预览

据《信息报》报道，OpenAI即将发布的最新模型GPT-5.6将一反常态，不会向公众开放。该公司计划仅在有限范围内向少量密切合作伙伴提供该模型，原因是特朗普政府要求其这样做。报道称，在本周的一次内部会议上，CEO萨姆·奥特曼向员工透露，在预览期内，政府将“逐客户审批访问权限”。奥特曼补充说，如果这次有限发布进展顺利，OpenAI希望在“几周后”实现更广泛的公开发布。

换言之，特朗普政府正施压OpenAI，使其采取与竞争对手Anthropic自愿采取的相同做法：将其最强大的AI模型“雪藏”起来。据《信息报》透露，OpenAI的新模型不仅正在接受政府审查，其员工还与政府在即将到来的发布中“紧密合作”。提出有限发布要求的机构包括国家网络总监办公室和科技政策办公室。

曾自诩对AI采取“不干预”政策的特朗普政府，近几个月来已转向推动联邦对新模型进行监管。本月早些时候，特朗普签署行政令，要求部分AI公司在公开发布新模型前，自愿提交给政府进行测试评估。

今年早些时候，Anthropic因其新前沿网络模型“Claude Mythos”仅通过“玻璃翼计划”向少数合作伙伴发布而引发争议。该公司辩称该模型“过于强大”，一旦落入不法之徒手中可能弊大于利。观察人士至今仍在争论，Anthropic的说辞究竟是营销噱头，还是防止强大模型被滥用的正当尝试——答案或许介于两者之间。

网络犯罪分子早已利用自动化工具作案，而在生成式AI时代，他们手中的数字武器比以往任何时候都多。大型语言模型已被证明擅长编写恶意软件，有些甚至能自主执行完整的勒索软件攻击。对于“Mythos”这类前沿网络工具，外界的具体担忧在于：它们理论上能以前所未有的速度发现并利用软件漏洞——这个速度是任何人类分析员都无法企及的。由于许多软件系统存在隐藏缺陷，成为企业网络的入侵通道，这对任何运行复杂软件基础设施的组织都构成了显而易见且重大的威胁。不过，由于这些模型仍不对公众开放，外界很难判断它们真实的威胁程度。

中文翻译：

OpenAI 即将发布其最新模型 GPT 5.6，据报道这次发布将与以往不同。据《The Information》报道，该公司不会向公众开放该模型，而是计划仅与少数精选的密切合作伙伴共享，原因是特朗普政府要求其如此。
在本周的一次会议上，CEO 萨姆·奥尔特曼据称告诉员工，在预览期内，政府将“逐一审批客户权限”。奥尔特曼还补充道，如果这次有限发布进展顺利，OpenAI 希望“几周后”能进行更广泛的大规模发布。
换句话说，特朗普政府似乎正在向 OpenAI 施压，要求其效仿 Anthropic 已主动采取的做法：将其最强大的 AI 模型秘而不宣。
根据《The Information》的消息，OpenAI 的新模型不仅正接受政府审查，其员工还在即将发布的版本上与政府“紧密合作”。据报道，要求进行有限发布的机构包括国家网络总监办公室和科技政策办公室。
特朗普政府最初宣称对 AI 采取“不干预”态度，但近几个月来却推动对新型模型实施联邦监管。本月早些时候，特朗普签署了一项行政命令，要求部分 AI 公司在公开发布新模型之前，自愿将其提交给政府进行测试和评估。
今年早些时候，Anthropic 宣布其新的前沿网络模型 Claude Mythos 将通过名为“Project Glasswing”的项目仅向一小部分合作伙伴开放，此举引发了不小的争议。Anthropic 辩称，该模型过于强大，如果落入不法之徒手中，可能弊大于利。此后，观察者们一直在争论 Anthropic 的这番说辞究竟是纯粹的营销噱头，还是为了保护强大模型不被滥用的合理尝试——答案或许介于两者之间。
网络犯罪分子使用自动化工具已有很长历史，但在生成式 AI 时代，他们如今拥有的数字武器比以往任何时候都要多。大语言模型已被证明善于编写恶意软件，有些甚至能够自主执行完整的勒索软件攻击。
像 Mythos 这样的前沿网络工具引发的具体担忧在于，它们表面上能够以人类分析师无法匹敌的速度识别和利用软件漏洞。由于许多软件系统都含有隐藏的漏洞，这些漏洞会成为企业网络的入口，因此这显然对任何运行复杂软件基础设施的组织构成了明显的重大威胁。话虽如此，由于这些模型仍未向公众开放，很难判断它们实际构成的威胁究竟有多大。

英文来源：

OpenAI’s release of its newest model, GPT 5.6, reportedly won’t be like its previous releases. Instead of distributing it to the public, the company plans to share it only with a select group of close partners because the Trump administration told it to, reports The Information.
At a meeting this week, CEO Sam Altman reportedly told staff that the government would be “approving access customer by customer” during a preview period. Altman reportedly added that if the limited release goes well, OpenAI hopes to follow with a general, broader release a “couple of weeks later.”
In other words, the Trump administration appears to be pressuring OpenAI to do what Anthropic is already voluntarily doing: keeping its most powerful AI models under wraps.
According to The Information, OpenAI’s new model is not only being reviewed by the administration, but its staffers also “worked closely” with the government on the upcoming release. The agencies that reportedly asked for a limited release were the Office of the National Cyber Director and the Office of Science and Technology Policy.
The Trump administration — which originally positioned itself as taking a “hands off” approach to AI — has in recent months pushed for federal oversight of new models. Earlier this month, Trump signed an executive order directing certain AI companies to voluntarily submit new models to the government for testing and evaluation before releasing them publicly.
Earlier this year, Anthropic sparked no small amount of controversy when it announced that its new frontier cyber model, Claude Mythos, would only be released to a small coterie of partners through a program called Project Glasswing. Anthropic argued that its model was simply too powerful and could, in the wrong hands, cause more harm than good. Observers have since debated whether Anthropic’s rhetoric is a mere marketing gimmick or a legitimate attempt to keep a powerful model from being misused. The answer may be somewhere in between.
Cybercriminals have used automated tools for a very long time, but in the age of generative AI, they now have more digital ammunition than ever before. LLMs have proven adept at writing malware, and some can even execute entire ransomware attacks autonomously.
The specific concern with frontier cyber tools like Mythos is that they are ostensibly capable of both identifying and exploiting software vulnerabilities at speeds that no human analyst could match. Since many software systems contain hidden bugs that act as entry points into enterprise networks, this obviously poses an obvious and significant problem for any organization running complex software infrastructure. That said, since these models remain closed to the public, it’s difficult to tell just how much of a threat they really are.