The White House Wants Anthropic to Block All Jailbreaks. That May Not Be Possible
The White House Wants Anthropic to Block All Jailbreaks. That May Not Be Possible
白宫要求 Anthropic 封堵所有“越狱”行为，但这或许根本无法实现
The Trump administration’s disagreement with Anthropic over its most advanced AI models appears to be fast coming to a head. 特朗普政府与 Anthropic 在其最先进 AI 模型上的分歧似乎正迅速走向白热化。
Trump officials tell Inner Loop that if Anthropic wants to rerelease Claude Fable 5, the AI model that they took offline with export controls last week over concerns about jailbreaking—a method of using prompts to get around a model’s safeguards—the company will need to take steps to actually address what the government alleges are vulnerabilities. 特朗普政府官员向《Inner Loop》透露，如果 Anthropic 想要重新发布 Claude Fable 5（该模型因“越狱”担忧于上周被出口管制措施强制下线，越狱是指通过提示词绕过模型安全防护机制的方法），该公司必须采取措施，切实解决政府所指出的漏洞。
Anthropic has said for days that the administration’s concerns are overblown and that the effects of the jailbreaks are minimal. It reiterated this position to the Commerce Department and the Office of the National Cyber Director, Sean Cairncross, in a technical meeting on Monday. Anthropic 多日来一直表示，政府的担忧被夸大了，且“越狱”带来的影响微乎其微。在周一的技术会议上，该公司向商务部及国家网络总监肖恩·凯恩克罗斯（Sean Cairncross）重申了这一立场。
But officials say they are past arguing whether the jailbreaks are significant, since the National Security Agency concluded that there are ways to disable guardrails on Fable 5, which are put in place to prevent users from accessing capabilities of the Mythos model related to cybersecurity, chemistry, and biology. 但官员们表示，他们已不再争论“越狱”是否严重，因为美国国家安全局（NSA）已得出结论：Fable 5 的安全护栏确实存在被禁用的可能，而这些护栏本是为了防止用户获取 Mythos 模型中涉及网络安全、化学和生物学方面的能力。
At this stage, the administration essentially views the situation as Anthropic’s problem to fix, according to three people familiar with discussions. 据三位知情人士透露，现阶段政府基本上将此视为 Anthropic 必须解决的问题。
Neither the Commerce Department’s Center for AI Standards and Innovation nor the National Security Agency has the staff or the bandwidth to be drawn into chasing down every conceivable jailbreak on every model that reaches the market, the people said. 知情人士称，无论是商务部下属的 AI 标准与创新中心，还是国家安全局，都没有足够的人手或精力去追踪市场上每一款模型中可能出现的每一个“越狱”漏洞。
As a result, the administration believes that Anthropic should be more proactive about continually testing not just Fable 5 but all of its frontier AI models to find potential jailbreaks and flag them to the government themselves. 因此，政府认为 Anthropic 应该更加主动，不仅要持续测试 Fable 5，还要对其所有前沿 AI 模型进行测试，以发现潜在的“越狱”漏洞并主动向政府报告。
But on a more fundamental level, it remains unclear how Anthropic is supposed to prevent jailbreaking. 但在更根本的层面上，Anthropic 究竟该如何防止“越狱”仍不明确。
Independent cybersecurity experts have increasingly taken the view that guardrails on AI models are only a stopgap solution, since skilled users and future AI models will find ways to bypass constraints—meaning that what the White House appears to want cannot be done. 越来越多的独立网络安全专家认为，AI 模型的安全护栏只是权宜之计，因为熟练的用户和未来的 AI 模型总能找到绕过限制的方法——这意味着白宫想要实现的目标可能根本无法达成。
A White House spokesperson declined to comment. 白宫发言人拒绝置评。
DNI = Do Not Invite
DNI（国家情报总监）= 请勿邀请
At the start of the week, Trump’s pick to serve as Acting Director of National Intelligence, Bill Pulte, was on track to never even start the job. Now, Trump has thrown him a lifeline—and it’s the permanent DNI nominee, Jay Clayton, who now faces the prospect of never serving in the role. 本周初，特朗普提名的国家情报代理总监人选比尔·普尔特（Bill Pulte）几乎注定无法上任。现在，特朗普向他伸出了援手，而正式的国家情报总监提名人杰伊·克莱顿（Jay Clayton）反而面临着可能永远无法担任该职位的局面。
To recap: Trump initially named Pulte, his housing finance chief, to replace outgoing DNI Tulsi Gabbard. 回顾一下：特朗普最初任命其住房金融主管普尔特接替即将离任的国家情报总监图尔西·加巴德（Tulsi Gabbard）。
Faced with bipartisan pushback because Pulte doesn’t have the national security experience required by law for the role and because he flagged allegedly questionable mortgage fraud accusations against Trump’s political enemies, Trump announced Clayton, the US attorney for the Southern District of New York, as his nominee for a permanent DNI. 由于普尔特缺乏法律规定的国家情报总监所需的国家安全经验，且他曾针对特朗普的政敌提出过涉嫌存疑的抵押贷款欺诈指控，此举遭到了两党的抵制。随后，特朗普宣布提名纽约南区联邦检察官杰伊·克莱顿为正式的国家情报总监。
Gabbard was scheduled to depart June 18, with Pulte’s first day set for June 19. But Senate Republicans wondered, if Clayton could have his hearing fast-tracked to June 17 and start by June 22, would Pulte even get into the building? 加巴德原定于 6 月 18 日离职，普尔特原定于 6 月 19 日上任。但参议院共和党人感到困惑：如果克莱顿的听证会能提前到 6 月 17 日并于 6 月 22 日上任，那么普尔特还有机会进入办公大楼吗？
On Wednesday, Trump blew up the plan. As part of a wider feud with Senate Republican leadership over the filibuster, Trump announced Clayton’s hearing would be delayed indefinitely, in an apparent effort to prevent Pulte from getting jumped. Senate Republicans then announced that the hearing would proceed, unless Clayton didn’t appear or his nomination was withdrawn. 周三，特朗普打破了这一计划。作为与参议院共和党领导层在“冗长辩论”（filibuster）问题上更广泛争端的一部分，特朗普宣布克莱顿的听证会将无限期推迟，此举显然是为了防止普尔特被“跳过”。参议院共和党人随后宣布，除非克莱顿不出席或其提名被撤回，否则听证会将照常进行。
The situation may be a body blow for the Office of the Director of National Intelligence, which Trump has directed Pulte to vastly downsize, and staffers have been unimpressed by what they see as Pulte’s minimal effort to get to know the agency and lack of regular briefings, people familiar with the matter said. 知情人士称，这种情况对国家情报总监办公室（ODNI）来说可能是一次重创。特朗普曾指示普尔特大幅精简该机构，而员工们对普尔特了解该机构的努力程度极低以及缺乏定期简报感到不满。
They have also sniped internally that Pulte appeared to them eager to get the perks of being DNI—the role comes with a security detail and often travel on government jets—without doing the often thankless work of producing sober intelligence briefings and corralling the nation’s intelligence agencies. 他们还在内部抱怨称，普尔特似乎只热衷于获得国家情报总监的特权（该职位配备安保团队，且经常乘坐政府专机），却不愿从事那些吃力不讨好的工作，比如撰写严谨的情报简报和协调全国的情报机构。
A White House spokesperson declined to comment, referring Inner Loop to Trump’s announcement on Truth Social. 白宫发言人拒绝置评，并将《Inner Loop》的问题转至特朗普在“真实社交”（Truth Social）上的声明。
Influence at the UFC
UFC 赛场上的影响力
Last week, Inner Loop reported on how White House aides expected a parade of donors and corporate executives to attend UFC Freedom 250 over the weekend in an effort to get close to Trump and senior officials from the administration. 上周，《Inner Loop》报道了白宫助手们如何预期大批捐赠者和企业高管将参加周末的 UFC Freedom 250 赛事，以寻求接近特朗普及其政府高级官员的机会。
The donors and executives duly showed up, both on fight night and at a number of parties held on the sidelines. 这些捐赠者和高管如期而至，不仅出现在比赛当晚，还参加了场外举办的多场派对。
Among the notable attendees was Paramount CEO David Ellison, who got approval from the Justice Department’s antitrust division last week to acquire Warner Bros Discovery. Also there was Meta CEO Mark Zuckerberg, who at one point was spotted talking to Trump. 出席的知名人士包括派拉蒙首席执行官大卫·埃里森（David Ellison），他上周刚获得司法部反垄断部门批准收购华纳兄弟探索公司。Meta 首席执行官马克·扎克伯格也出席了活动，并一度被目击与特朗普交谈。
Earlier in the weekend, Meta hosted a private party at the Ned’s Club overlooking the White House, attended by a number of Trump officials including the acting attorney general Todd Blanche, White House deputy chief of staff James Blair, White House press secretary Karoline Leavitt, and interior secretary Doug Burgum. 周末早些时候，Meta 在俯瞰白宫的 Ned’s Club 举办了一场私人派对，多位特朗普政府官员出席，包括代理司法部长托德·布兰奇（Todd Blanche）、白宫副幕僚长詹姆斯·布莱尔（James Blair）、白宫新闻秘书卡罗琳·莱维特（Karoline Leavitt）以及内政部长道格·伯格姆（Doug Burgum）。
Also there were some Trump family members—Jared Kushner and Ivanka Trump, as well as Kai Trump—who mingled with other guests including former Clinton strategist Adrienne Elrod, Axios cofounders Mike Allen and Jim VandeHei, Fox News host Shannon Bream, Washington AI network founder Tammy Haddad, and former Trump strategist Kellyanne Conway. 出席的还有一些特朗普家族成员——贾里德·库什纳（Jared Kushner）、伊万卡·特朗普（Ivanka Trump）以及凯·特朗普（Kai Trump）。他们与包括前克林顿竞选策略师阿德里安·埃尔罗德（Adrienne Elrod）、Axios 联合创始人迈克·艾伦（Mike Allen）和吉姆·范德海（Jim VandeHei）、福克斯新闻主持人香农·布里姆（Shannon Bream）、华盛顿 AI 网络创始人塔米·哈达德（Tammy Haddad）以及前特朗普策略师凯莉安·康威（Kellyanne Conway）在内的其他嘉宾进行了交流。
This is an edition of Hugo Lowell’s Inner Loop newsletter. Read previous newsletters here. 这是雨果·洛厄尔（Hugo Lowell）的《Inner Loop》通讯。点击此处阅读往期内容。