政府如何判定 OpenAI 的前沿模型可以安全发布？

OpenAI 正向公众广泛推出其最新的先进大语言模型 Sol。Sol 被认为至少与 Anthropic 的 Fable 不相上下，而后者的能力（或所有权）曾让白宫倍感压力，甚至一度被禁止公开访问。

那么，这些模型是如何获得发布许可的呢？简短的回答是：没人真正清楚。

“坦白说，我无法了解那些具体的流程，所以是的，我确实觉得自己掌握的信息不足以判断它们是否充分，”乔治城大学安全与新兴技术中心（CSET）的高级研究分析师 Mina Narayanan 告诉 TechCrunch。“Anthropic 的确表示他们一直在与政府对话，并且开发了一个检测越狱企图的分类器，还实施了纵深防御策略来防止将来的越狱，但政府与 Anthropic 和 OpenAI 之间的对话究竟是什么样子，外界并不清楚。”

前特朗普政策顾问、现供职于 OpenAI 的 Dean W. Ball 上个月在其新闻通讯中撰文称，“没人知道获得许可的要求是什么”。

曾联合创立 Databricks、Perplexity 和 Laude Institute 的计算机科学家 Andy Konwinski 表示，他从未和任何了解该流程的人交谈过，即便是前沿实验室的员工也不例外。“这是一个关乎存亡的问题，”他对 TechCrunch 说。“不管是否关乎安全，关键在于谁拥有决策的权力——谁在把关，谁来决定许可？”

特朗普政府上台18个月后，关于如何推进仍不明朗，尽管——或者正如一些批评者所指出的，正是由于——制定政策的正是业内人士。上个月，在数周的内斗之后，一项行政令公布了一份评估前沿模型的路线图，但除了明确不会设立什么之外，具体细节仍有待充实。“不会有 AI 领域的 FDA（食品药品监督管理局），”曾担任 Andreessen Horowitz 合伙人、直到上月还在白宫担任 AI 高级顾问的 Sriram Krishnan 告诉《金融时报》。

值得注意的是，对于哪些类型的模型需要接受政府审查，以及应由哪个或哪些机构执行这些评估，目前仍未达成一致。眼下，商务部的人工智能标准与创新中心似乎正牵头负责，但行政令要求六个内阁机构在八月初之前确定最终流程。在此期间出现的情况，充其量只是临时的、权宜之计。

OpenAI 首席执行官 Sam Altman 在 CNBC 上表示，这一流程包含了与商务部长 Howard Lutnick、财政部长 Scott Bessent 和美国国家网络总监 Sean Cairncross 等官员的对话，但并不清楚测试这些模型的专家是谁，以及他们是如何测试的。OpenAI 拒绝与 TechCrunch 分享政府流程的细节，但指向了该模型安全卡中由英国 AISI、SecureBio 和 Irregular 等组织完成的多项外部评估结果。

与 Anthropic 推出 Fable 时类似，OpenAI 在更广泛发布之前将模型预览版提供给了政府和特定用户，但我们不知道这些用户都有谁，也不清楚他们是如何被选中的。在六月下旬的一篇博客文章中，该公司表示“我们认为这类政府准入流程不应成为长期的默认方式”，并称将和政府合作探索一条不同的前进路径。

然而，这些对话的背景包括：据报 Altman 提议将 OpenAI 最高 5% 的股权捐赠给政府的所谓“特朗普账户”，以及 OpenAI 总裁 Greg Brockman 作为特朗普中期选举政治运作已知的最大公开捐助者的角色。外部观察人士很难将这些活动与政府对 Sol 明显较为宽松的监管方式割裂开来。

另一方面，Anthropic 的 Fable 曾因美国政府禁止外国公民使用而短暂撤出广泛访问，部分原因是确实担心用户通过越狱模型获取黑客能力，部分原因也在于 Anthropic 与特朗普政府之间的性格冲突。出口禁令的威胁可能也促使 OpenAI 对政府的（未知）要求表现得更加配合。

从行业角度看，宽松的监管方式或许很好，但一种依赖与行政官员私人关系的监管，却制造了不确定性和不良激励。

Konwinski 告诉 TechCrunch，他担心真正的技术专家——“安全研究人员、对齐研究人员、可解释性研究人员，也包括数据人员和来自整个技术栈的人员”——在模型发布流程中发挥的作用不够。

Konwinski 主张，“开放共享”是真正平衡安全与创新的最佳途径。他列举了 FDA、NIH 或国家实验室等模式，这些机构会召集研究人员、政府官员和私营企业，就安全问题达成共识。

这在某种程度上归结为激励 AI 研究人员十多年的资本主义动机，并在 Elon Musk 挑战 OpenAI 公司结构的诉讼中在法庭上展现出来。Ball 指出，AI 业务的特性要求各家公司在模型发布并领先竞争对手后不久就收回大部分训练成本。

“即便他们的意图是好的，运营程序中本身就内置了非常清晰的法律义务和信托责任，”Konwinski 指出。

Ball 在他的帖子中主张，前进的道路将有赖于政府授权的第三方审计机构，它们将评估前沿实验室的安全方法。Konwinski 同样看好聚焦研究组织等新机构形式，它们可以帮助学术界和非营利界更多无利害关系的专家接触并评估前沿模型。

就目前而言，围绕 AI 开发的秘密不会消失，但也将为这个美国民众日益持怀疑态度的行业埋下政治挑战。“人们并不觉得是负责任的人在推动这些变革，”威斯康星大学麦迪逊分校计算机科学教授 Remzi Arpaci-Dusseau 上周在 Open Frontier 会议上说。

在同一活动中，创办了最成功的量化对冲基金之一 Two Sigma 的计算机科学家 David Siegel 请与会者“想象一种我认为非常糟糕的情形：少数几家公司控制着技术；政府在其秘密实验室中评估该技术是否适宜使用；而公众和科学界对这些东西几乎毫无接触。”

看来我们不需要想象了。

OpenAI is rolling out its latest advanced LLM, Sol, for wide public access. Sol is considered to be at least on par with Anthropic’s Fable, a model whose capabilities (or ownership) stressed out the White House enough that it was briefly banned from public access.

So how did these models get the OK for release? Short answer: Nobody’s quite sure.

“Frankly, I don’t have visibility into those exact processes, so yes, I don’t feel like I have enough information to say whether they’re adequate or not,” Mina Narayanan, a senior research analyst at Georgetown’s Center for Security and Emerging Technology, told TechCrunch. “Anthropic did say that they were in conversations with the government, and that they developed a classifier to detect jailbreak attempts, and they’ve implemented defense-in-depth strategies to prevent future jailbreaks, but exactly what that dialog looked like between the government and Anthropic and OpenAI is unclear.”

Dean W. Ball, a former Trump policy advisor who now works for OpenAI, wrote that “nobody knows what the requirements are to get licensed” in his newsletter last month.

Andy Konwinski, a computer scientist who co-founded Databricks, Perplexity, and the Laude Institute, said he’s never spoken to anyone who understands the process, even employees at frontier labs. “It’s existentially a problem,” he tells TechCrunch. “Safety or not, it’s about who has the power to make decisions — who gatekeeps and decides on permissions?”

Eighteen months into the Trump administration, there is still little clarity about how to move forward, despite — or, some critics allege, because — of the industry figures setting policy. Last month, after weeks of infighting, an executive order was published laying out a roadmap for evaluating frontier models, but the specifics have yet to be filled in, other than what won’t exist. “There will not be an FDA for AI,” Sriram Krishnan, a former Andreessen Horowitz partner who served as a senior advisor for AI in the White House until last month, told Financial Times.

Notably, there’s still no agreement on what kinds of models require government scrutiny, or what agency or agencies should perform those evaluations. For now, the Department of Commerce’s Center for AI Standards and Innovation seems to be taking the lead, but the executive order instructs six cabinet agencies to determine a final process by early August. What has emerged in the meantime is, at best, ad hoc.

OpenAI CEO Sam Altman said on CNBC that the process involved conversations with the officials like Secretary of Commerce Howard Lutnick, Secretary of the Treasury Scott Bessent, and U.S. national cyber director Sean Cairncross, but it’s not clear who the experts that tested the models were or how they did that. OpenAI declined to share details on the government’s process with TechCrunch, but pointed to the results of several external evaluations by organizations like U.K. AISI, SecureBio, and Irregular in the latest model’s safety card.

As with Anthropic’s Fable roll-out, OpenAI previewed the model for the government and select users ahead of wider release, but we don’t know who all of those users were or how they were chosen. In a late June blog post, the company said “we don’t believe this kind of government access process should become the long-term default,” saying it would work with the government to develop a different path forward.

The backdrop to those conversations, however, includes Altman reportedly offering as much as 5% to OpenAI’s equity for the administration’s so-called “Trump Accounts,” and OpenAI president Greg Brockman’s role as the largest publicly known donor to Trump’s mid-term political operation. It’s hard for outside observers to separate those activities from the government’s apparently lighter-touch approach to regulating Sol.

Anthropic’s Fable, on the other hand, was briefly pulled from wider access when the U.S. government forbade its use by foreign nationals, partly because of real concerns about users jail-breaking the model to access hacking capabilities and partly due to personality clashes between Anthropic and the Trump administration. The threat of an export ban may have also led OpenAI to be more cooperative with the government’s (unknown) requests.

From an industry perspective, a hands-off approach to regulation might be nice, but one that depends on personal connections to administration officials creates uncertainty and bad incentives.

Konwinski told TechCrunch that he worries true experts in this technology — “safety researchers, alignment researchers, interpretability researchers, but also data people, and people from all over the stack” — aren’t playing enough of a role in the model release process.

Konwinski argues that an “open commons” is the best way to actually balance safety and innovation. He points to models like the FDA, the NIH, or the national labs, which convene researchers, government officials, and private companies to reach a consensus on safety issues.

Some of that comes down to the incentives of capitalism that have motivated AI researchers for more than a decade, and played out in the court room during Elon Musk’s lawsuit challenging OpenAI’s corporate structure. Ball points out that the nature of the AI business requires companies to recoup much of their training costs shortly after their models are released and are further ahead of the competition.

“Even if their intentions are good, there’s very clear legal obligations and fiduciary responsibility that are built right into the operating procedures,” Konwinski points out.

Ball, in his post, argued that the way forward will depend on third-party auditing organizations, licensed by the government, that will evaluate frontier labs’ approach to safety. Konwinski, too, is bullish about new institutional formats like focused research organizations that could help more disinterested experts from academia and the nonprofit world access and evaluate frontier models.

For now, the secrecy around the development of AI isn’t going away, but it also will seed political challenges for an industry that Americans increasingly view with skepticism. “There’s not a sense that responsible people are driving forward these changes,” University of Wisconsin-Madison computer science professor Remzi Arpaci-Dusseau said last week at the Open Frontier conference.

At the same event, David Siegel, the computer scientist who founded Two Sigma, one of the most successful quantitative hedge funds, asked attendees to “imagine a situation, which I think would be very bad, [where] a small number of firms control the technology; the government, in their secretive laboratories, is evaluating whether or not the technology is suitable for use; and the general public and scientific community doesn’t really have any access to any of that stuff.”

It seems like we don’t need to imagine it.