网易首页 > 网易号 > 正文 申请入驻

白宫押注“超级智能”,撞上华盛顿的新权力方程式

0
分享至

华盛顿在人工智能政策处理方式上正经历一场剧烈转变。过去一个月里,白宫宣布成立一个新的特别工作组,负责监督奔向超级智能;的商业竞赛。几周后,美国人走向投票站,在地方、州和联邦各级选举中以压倒性优势选出民主党候选人。可新当选的议员们还没来得及庆祝完,前英伟达首席执行官黄仁勋的婚外情消息抢走了全部注意力,成了全城最热的话题。

在白宫努力设法防止美国人工智能模型构成国家安全威胁又不向外国竞争对手让出优势之时,其议程的时间可能正在流逝。中期选举的结果,让特朗普总统在第二任期内第一次面对一个对立的国会——民主党人从共和党手中同时夺走了参议院和众议院的控制权。过去几周,两党内部的对华强硬派;也在国会与联邦机构之间拿到了可观的谈判筹码。

在民主党人明年一月正式接掌之前,参议员们还得先扛住通过年度国防授权法案的压力。总统提出的逾 1 万亿美元国防预算被参议院民主党人指为臃肿的军费开支;。就在参议院围绕这份庞杂的立法包裹吵得不可开交之际,两党议员普遍预计,鉴于英伟达近来的是非,收紧对华先进芯片出口的条款会被塞进这部法案。

多年来零敲碎打的政策制定让华盛顿政界此刻感受到一种不同寻常的紧迫感:多个政府部门正对人工智能采取更多干预。9 月下旬,白宫召集六位人工智能行业领袖签署了一项自愿性人工智能安全协议;不到一周后,白宫又任命了一个特别工作组,其任务是评估关键领域的 AI 风险、建立向政府报告安全事件的规程,并梳理政府可用于监管人工智能行业的现有联邦权限。这个工作组成立以来已开过两次会,议题包括重新授权一项责任豁免条款(该条款允许企业与政府共享网络安全漏洞信息),以及为整个人工智能行业确立统一的行为准则。

北京的观察人士对于这一正在成形的战略或许并不陌生。2017 年,中国将少数几家科技企业指定为国家人工智能领军企业,力图让商业雄心与国家经济优先方向保持一致。近十年过去,在技术历经多次跃升后,白宫正朝相似的方向迈进。特朗普政府向来照顾美国商界领袖的利益,如今必须引导实力雄厚的私营企业和联邦机构朝其宣称的快速创新议程前进。但除非启用紧急权力,否则它会在不集中掌控本国工业根基的前提下尝试实现该目标。

特朗普政府的相关计划还需要和国会博弈,后者可能很快要求在美国 AI 竞争的监管事宜中拥有更多话语权。中期选举使得民主党掌控参众两院多数席位,国会大概率将在 2027 年大力推动 AI 监管,但立法未必能够落地,总统仍可否决国会法案。

不过这有一个前提:国会共和党人不会出于党派博弈考量,出手阻滞立法推进。华盛顿非营利智库信息技术与创新基金会(ITIF)部分经费来自科技公司,该机构主席丹尼尔·卡斯特罗称:在一些议题上两党有着共识,比如保障普通用电用户不必承担数据中心带来的公用事业开支。但两党同样心照不宣:没人愿意成全对方拿下政策成果。;

AI 安全倡导者的态度要乐观得多。无党派倡导组织美国人促进负责任创新;致力于防范 AI 危害,其传讯副总裁克里斯·麦肯齐称:我非常倾向于新一届国会将推出刚性明确的 AI 监管法规,而不是继续依赖企业自愿承诺的模式。;

前沿 AI 研发企业正身处变幻动荡的政策环境之中,它们愈发认定,联邦监管避无可避,眼下正竞相争取话语权,企图左右监管规则的走向。

眼下 AI 行业面临的最直接压力来自华盛顿的联邦机构,这些机构的官员已经表明态度:愿意把出口管制当作缰绳,沿着整个 AI 技术栈层层收紧,以此管束美国企业。

长期以来,华盛顿主要依靠出口管制,调控流向中国科技竞争对手的高端芯片供应量。今年 6 月,研究人员查出模型存在安全漏洞,美国商务部随即援引《2018 年出口管制改革法案》下达紧急指令,一夜封禁境外对 Anthropic 两款顶尖模型的访问通道。这道命令前后执行了 18 天;在此期间,商务部与 Anthropic 达成协议,随后解除了禁令。身兼五角大楼高级官员、白宫全新 AI 超级智能专项工作组成员的埃米尔·迈克尔称,这起事件是一记警告,面向所有不愿配合联邦政府的前沿 AI 实验室。

在 ITIF 的卡斯特罗看来,今夏这场 Anthropic 风波,已经给 AI 全行业敲了一记警钟:切勿对抗特朗普政府的决定。业内从中得到的启示很明确:只要白宫下令不得推出某款模型,或是相关事项存疑,企业就不能把模型对外发布。;

国商务部周一再度援引《2018 年出口管制改革法》中赋予的权限,发布了一份长达 132 页的临时最终规则,意在阻止美国的境外对手向美国企业租用或远程调用先进算力。这也是联邦政府针对英伟达相关事件落地的首项实质性举措。该规则标注为 临时最终规则;,意味着政府可以跳过常规的公告公示与公众意见征询流程。负责执行该规则的商务部下属工业与安全局(BIS)在新闻稿中称,这份规则由财政部与白宫国家安全委员会牵头起草。

在此新规出台之前,财政部长斯科特・贝森特曾在一档公开新闻节目上点名批评由霍华德・卢特尼克执掌的商务部。贝森特指出,尽管其他联邦机构早已提出警示,商务部仍放任向 Megaspeed 这类企业持续进行出售高端芯片的交易。卢特尼克因对英伟达以及黄仁勋采取宽松监管态度,正承受各方审视。本周晚些时候,他将随同特朗普总统赴中国出席亚太经合组织(APEC)峰会。

受到警示的不只是美国本土研发方。10 月,美国国土安全部下令联邦民用机构从电脑系统中卸载深度求索(DeepSeek)相关产品,禁令范围进一步扩大至所有依托深度求索技术开发的第三方产品与服务,这也迫使美国政府承包商重新评估是否继续使用中国开源权重模型。

随着黄仁勋相关事件接连曝光,美国与亚洲商业伙伴之间原本轻松的往来氛围已然降温。特朗普总统及其核心顾问认为,黄仁勋的行为并不构成叛国,但国会议员并不认同。一些民主党参议员已致函白宫、国务院与商务部,要求留存相关记录,内容覆盖英伟达向中国及关联实体供应芯片的全部信息,同时呼吁出台更严格的芯片出口管制,并对英伟达与黄仁勋展开调查。

然而,国会正在纠缠一个比英伟达更大的问题:要不要为前沿 AI 研发建立长效政府监管框架。这场辩论的核心问题表面上简单,实则暗藏博弈:如果政府在安全测试环节检出一款 AI 模型存在危险能力,是否应当拥有叫停其公开发布的权力?

得州共和党参议员、参议院商业委员会现任主席特德·克鲁兹始终反对授予联邦监管当局权限,使其可以仅凭潜在风险就阻断消费者使用 AI 模型。他坚持认为,诉讼本身带来的威慑足以倒逼企业保证 AI 产品对消费者安全。哪家公司推出的产品酿成重大损害,都会被官司诉到彻底垮台。;克鲁兹 9 月说。

中期选举尘埃落定,克鲁兹参议员一下子成了得州仅剩的、保有全州公职的共和党人。等到明年 1 月,民主党将拿下参众两院控制权开启新一届国会,克鲁兹预计会把参议院商业委员会主席的席位交给该委员会的民主党资深同僚——来自华盛顿州的玛丽亚·坎特韦尔参议员。针对共和党全国选战失利,克鲁兹在 X 上发声,进一步重申自身经济立场:我们要重塑为真正代表蓝领阶层的政党,这条路不靠拥抱社会主义,而是做秉持纯粹自由市场理念的保守派。;

与此同时,坎特韦尔参议员曾担任参议院商业委员会主席,如今又有望接替克鲁兹执掌该委员会。她在 10 月抛出了一套六模块前沿 AI 监管框架,主张将前置测试制度立法固化,由政府科研人员在商用 AI 模型对外发布前完成独立评估。这套框架不能和众议院的《前沿法案》(FRONTIER Act)混为一谈,尽管二者聚焦相近议题,但外界普遍预计,明年会有约束力更强的新方案取代这份众议院法案。

OpenAI 全球事务主管克里斯·莱哈内将《前沿法案》评价为不错的起步框架;,并呼吁对 AI 模型实行由政府强制要求的发布前审计。不过,前沿 AI 企业的政策偏好正碰上一个越来越怀疑的国会:Anthropic 代表上个月向国会推介了自研的 AI 安全法案草案,不料引来 AI 安全倡议人士的集体反对。外界看到外泄草案内容后,指责这份文本优先袒护 AI 实验室,而非把消费者保护放在首位。全美多家民权组织也向联邦议员划下红线:他们不会接受任何全国性 AI 安全法案——如果它会削弱纽约、伊利诺伊等多个州已经落地的专项 AI 保护规则。

中期选举让 AI 政策博弈的政治底色发生了转向。由硅谷一众知名投资人、企业高管支持的亲 AI 创新超级政治行动委员会引领未来;,在选前数月没有任何新增捐款,而倡导收紧 AI 安全约束的同类 PAC 却风头日盛。时至 10 月,引领未来;的竞选储备金仍有 2000 多万美元,但在舆论哗然之后,主要出资者已经开始和该委员会保持距离。

我认为白宫此前的预判是,AI 议题很难打动普通选民,但对竞选捐资人而言至关重要,;ARI 的麦肯齐说。但这一回,执政当局很可能算错了。公众其实十分关心 AI,这场选举里我们已经感受到相关的民意反弹,;麦肯齐补充道。他所在机构追踪发现,参选候选人围绕 AI 政策议题,收获了前所未有的选民热情与参与度。

特朗普政府正试图摸索一套行之有效的 AI 竞争策略,但这些尝试几乎无法让中国方面相信:美国的技术管制会随之放宽。所有调整方向看上去都是在进一步收紧管控,;ITIF 的卡斯特罗说,没人愿意落一个在国家安全议题上态度偏软的名声。;

这为本周在深圳举行的亚太经合组织(APEC)峰会铺就了一层微妙的底色。美国政府有强烈意愿推动机构投资,防止中美贸易关系继续走坏。但白宫能否兑现双边 AI 相关政策目标,取决于中期选举尘埃落定之后,它还能不能对国会以及 AI 产业保有把控力。

白宫的超级智能专项工作组定于明年 1 月提交研究结论,而新一届国会也将在同一节点着手推进自身的政策议程。这项工作组的人员架构已经可能迎来变动,特朗普总统用移民强硬派斯蒂芬·米勒替换了白宫幕僚长苏西·怀尔斯——怀尔斯原本也是该专项工作组的成员。截至规定问询时限,白宫并未作出回应,确认米勒是否会接替怀尔斯加入这一工作组。

如果政府希望打造一套独属于美国的应对思路,来抗衡中国龙头企业的国家扶持战略,它要跨过的第一道关卡是一套新的华盛顿权力方程式:在这套刻意分权的设计之下,无论是白宫,还是 AI 产业本身,都已无法再指望国会与联邦机构无条件为自己的政策议程背书。

本周出炉的选举结果说明,全美选民,无论身处南部腹地还是北方地区,都愿意跳出数十年以来的投票范式,投票选出承诺优先回应消费者诉求的议员。此前两年,白宫一路推行加速发展的 AI 政策,国会鲜少掣肘;待到明年,这批新晋参众议员入驻华盛顿,很可能倒逼 AI 行业收敛步伐、更加谨慎行事。(作者|Mohar Chatterjee,编辑|李婧滢)

英文原文如下(ORIGINAL):

The White House’s gamble on superintelligence; meets a new power equation in DC

Washington is going through a seismic shift in how it handles AI policy. In the past month, the White House announced a new taskforce dedicated to overseeing the commercial sprint towards superintelligence.; A few weeks later, Americans headed to the polls to elect an overwhelming sweep of Democratic Party candidates to local, state and federal offices. But even before the new Congress-elects were done celebrating, news of former Nvidia CEO Jensen Huang’s extramarital affairs seized public attention, quickly becoming the talk of the town.

As the White House grapples with ways to prevent American AI models from posing national security threats without ceding ground to foreign rivals, the clock might be ticking down on their agenda. The midterm election results have left President Trump facing an adversarial Congress for the first time during his second term, with Democrats wresting control of both the Senate and House from Republicans in the midterm elections. And in the past few weeks, China hawks in both parties have also gained valuable negotiating ground across Congress and federal agencies.

Before Democrats officially take the wheel in January, Senate lawmakers are under pressure to approve the annual defense bill. The President’s request for a national defense budget of over $1 trillion has drawn sharp criticism from Senate Democrats for being bloated defense spending. As the Senate hotly debates this mammoth legislative package, lawmakers across the aisle expect measures to tighten advanced chip exports to China to be a narrow area of agreement, in light of the recent news about Nvidia.

Years of piecemeal policymaking have led to a moment of unusual urgency within the beltway, as multiple branches of government embrace a more interventionist approach to artificial intelligence. In late September, the White House assembled six AI industry leaders to sign a voluntary AI safety accord, following up less than a week later with the appointment of a task force whose mission is to evaluate AI risks in critical sectors, establish protocols for reporting safety incidents to the government, and identify existing federal authorities the government can use to regulate the AI industry. The task force has already met twice since its formation, most recently discussing the reauthorization of a liability shield that allows businesses to share information about cybersecurity vulnerabilities with the government, as well as establishing a common standard of conduct for the AI industry writ large.

The emerging strategy may sound familiar to observers in Beijing. In 2017, China designated a handful of technology companies as national AI champions, seeking to align commercial ambitions with the country’s economic priorities. Nearly a decade and multiple leaps in technological progress later, the White House is moving in a similar direction. The Trump administration — which has a track record for favoring the interests of American business leaders — must now steer powerful private companies and federal agencies towards its stated agenda of rapid innovation. But it will attempt to do so without any centralized authority over the country’s industrial base unless it invokes emergency powers.

The Trump administration’s ambitions will need to contend with a Congress that may soon demand greater say in how America’s AI race is governed. With a decisive majority in both the House and Senate after the midterm elections, Democrats in Congress will almost certainly enter 2027 with greater appetite to dive into AI regulation— though not necessarily with greater success, since the President still has the power to veto any bills Congress passes.

And that’s presuming Congress Republicans don’t stall legislative momentum based on partisan calculus. There’s a lot of bipartisan consensus on certain issues like protecting ratepayers from the utility costs of data centers,; said Daniel Castro, president of the Information Technology and Innovation Foundation, a nonprofit think tank based in Washington that is funded in part by tech companies. But there’s also a sense that neither side wants to give the other side a win.;

AI safety advocates are more hopeful. I’m very optimistic that we’re going to see the next Congress act on hard and fast AI regulations rather than continuing down the path of voluntary commitments,; said Chris MacKenzie, vice president of communications at Americans for Responsible Innovation, a nonpartisan advocacy group focused on mitigating AI harms.

It’s a tumultuous policy landscape for frontier developers, who increasingly regard some federal oversight as inevitable and are now competing to shape its contours.

The most immediate pressure on the AI industry is coming from Washington’s federal agencies, whose officials have demonstrated willingness to wield export controls across

For years, Washington used export controls primarily as a way to titrate the supply of advanced chips to Chinese technological competitors. In June, the Commerce Department cut off foreign access to two of Anthropic’s most advanced models overnight through an emergency directive issued under the Export Control Reform Act of 2018, after researchers discovered security vulnerabilities in the models. The agency lifted the order 18 days after issuing it, once it had reached an agreement with Anthropic. Emil Michael, a senior Pentagon official who is also on the White House’s new AI superintelligence task force, called that episode a warning to any frontier lab that doesn’t cooperate with the federal government.

According to ITIF’s Castro, the Anthropic incident this summer sent a clear message to the AI industry: do not resist the Trump administration’s will. The takeaway for industry was that if [the White House] tells you not to release a model, or if there are any doubts — you’re gonna need to not release the model,; Castro said.

The Commerce Department used its authority under the Export Control Reform Act of 2018 again on Monday to issue a new 132-page interim final rule; targeted at preventing the U.S. foreign rivals from leasing or remotely accessing advanced computing power from American companies, marking the federal government’s first concrete action in response to the news about Nvidia. The rule’s interim final designation; allows the government to bypass the usual public notice and comment periods. The Bureau of Industry and Security — the office within the Commerce Department responsible for enforcing this rule — said in a press release that the architects of the rule are the Treasury Department and the White House National Security Council.

The new rule came after Treasury Secretary Scott Bessent held the Commerce Department — headed by Secretary Howard Lutnick — at fault on a public news program for allowing the sale of advanced chips to companies like Megaspeed to continue despite concerns expressed by other federal agencies. Commerce Secretary Lutnick, who is facing scrutiny over his lenient treatment of Nvidia and Jensen Huang, is scheduled to join President Trump later this week for the Asia-Pacific Economic Cooperation (APEC) summit in China.

It’s not just domestic companies on notice. Back in October, the Department of Homeland Security ordered federal civilian agencies to remove DeepSeek products from their computer systems, extending the ban to any third-party products and services developed using DeepSeek, forcing American government contractors to reconsider any use of Chinese open-weight models.

There is certainly a chill enveloping casual relationships with Asia-based business partners right now, after the recent revelations about Jensen Huang’s conduct. While President Trump and some of his key advisors have expressed that they don’t view Huang’s actions as treasonous, Congress lawmakers are less convinced. A few Democratic senators have asked the White House, State and Commerce Departments to preserve records pertaining to Nvidia’s supply of chips to China and Chinese-affiliated entities, joining other Democrats in calling for stricter chip export controls and probes into Nvidia and Huang.

Congress, however, is grappling with a question even bigger than Nvidia: whether to establish more permanent government oversight on frontier AI development. At the center of the debate is a deceptively simple question: If the government discovers that an AI model has dangerous capabilities during a safety test, should it have the power to prevent the model’s public release?

Senator Ted Cruz, the Texas Republican who currently chairs Senate Commerce Committee, has resisted giving federal regulators the power to block off model access to consumers based on the potential to cause harm, insisting that the threat of legal action is enough incentive for companies to ensure AI products are safe for consumers. If any company sells a product that causes massive harm, they’ll get sued into oblivion,; Sen. Cruz posted on X in September.

After the midterm elections, Sen. Cruz suddenly found himself to be the only Republican remaining in a statewide political seat in Texas. Come January, when a new Congress convenes under Democrat leadership, Cruz is expected to hand over the chair’s gavel to his Democratic counterpart on the Commerce Committee, Senator Maria Cantwell of Washington. Cruz reacted to his party’s nationwide election losses on X by doubling down on his economic stances: We will rebuild as a true blue-collar party, not by embracing socialism, but by being real free-market conservatives.;

Meanwhile, Sen. Cantwell, who is both his predecessor and likely successor as Senate Commerce Chair, introduced a six-part framework to govern frontier AI development in October. She wants to codify a testing regime that will allow government scientists to independently evaluate commercial AI models before public release. Cantwell’s framework should not be confused with the FRONTIER Act introduced in the House which deals with similar topics, but is widely expected to be replaced with a stronger proposal next year.

OpenAI’s chief global affairs officer Chris Lehane called the FRONTIER Act a strong starting point; and has already endorsed the government-mandated, pre-release audits of AI models. Though the policy preferences of frontier AI developers are landing on an increasingly skeptical Congress: Anthropic representatives pitched their own AI safety bill to Congress in the past month, but drew widespread backlash from AI safety advocates who criticized the draft bill’s leaked text for shielding AI labs over protecting consumers. Civil rights groups across the country have already made it clear to federal lawmakers that they will not accept a national AI safety bill that could weaken specific AI-related protections already adopted by multiple state governments, including in New York and Illinois.

The midterm elections have revealed shifting a political ground underneath the AI policy negotiations. Leading the Future, a pro-AI innovation super PAC backed by prominent Silicon Valley investors and executives, reported no new donations in the months leading up to the midterms, even as PACs campaigning for stronger AI safeguards gained prominence. As of October, Leading the Future still had over $20 million left in their war-chest, but some major donors had already distanced themselves from the super PAC after public outcry.

I think the White House bet was that AI wouldn’t actually matter to American voters, but it matters a lot to [political campaign] donors,; ARI’s MacKenzie said. But the incumbent administration might have miscalculated this time. The public does care about AI, and we’ve seen the backlash this election,; MacKenzie said, noting that his organization had tracked an unprecedented level of voter enthusiasm and engagement on AI policy matters among candidates campaigning for public office.

The administration’s attempts to figure out a winning AI strategy offer little assurance to China that American technology restrictions will ease. All the adjustments are about making things tighter,; ITIF’s Castro said. Nobody wants to be seen as being soft on national security.;

That leaves an awkward backdrop for the APEC summit in Shenzhen later this week. The administration has every incentive to encourage institutional investments and avoid any deterioration in U.S.-China trade relations. But the White House’s capacity to deliver on bilateral AI policy targets will depend on whether the administration emerges with any control over Congress and the AI industry once the dust has settled from the midterms.

The White House’s superintelligence task force is due to deliver its findings in January, just as a newly elected Congress prepares to assert its priorities. Already the task force roster might be shifting, after President Trump replaced Susie Wiles (who was also on the superintelligence task force) with immigration hardliner Stephen Miller as the White House chief of staff. The White House did not respond by a deadline to confirm whether Miller would replace Wiles on the task force as well.

If the administration is hoping to mount a distinctly American answer to China’s national champions strategy, its first challenge will be navigating a new power equation in Washington where neither the White House nor the AI industry is likely to have the final say.

This week’s election results have demonstrated that American voters across the country — from the deep South to the cold North — are willing to buck decades of voting trends to elect representatives who promise to put consumer concerns first. After two years of accelerationist policies from the White House and little pushback from Congress, a new class of Senators and House representatives headed to Washington next year might force the AI industry to tread more carefully than before.

特别声明:以上内容(如有图片或视频亦包括在内)为自媒体平台“网易号”用户上传并发布,本平台仅提供信息存储服务。

Notice: The content above (including the pictures and videos if any) is uploaded and posted by a user of NetEase Hao, which is a social media platform and only provides information storage services.

相关推荐
热点推荐
54年来头一遭,打破美国总统常规,特朗普在深圳究竟看到了什么?

54年来头一遭,打破美国总统常规,特朗普在深圳究竟看到了什么?

阿天爱旅行
2026-11-21 06:15:27
妻子不愿办身份证,江西男子2021年才得知,她是越狱27年的人贩子

妻子不愿办身份证,江西男子2021年才得知,她是越狱27年的人贩子

夜夜思奇语
2026-11-20 15:00:26
哈登轰51分创11项神迹:一数据历史第一+超越科比+骑士生涯新高

哈登轰51分创11项神迹:一数据历史第一+超越科比+骑士生涯新高

Emily说个球
2026-11-21 10:57:46
多国首脑在深圳闲逛,每个人路线都很有意思,普京发出了内心所想

多国首脑在深圳闲逛,每个人路线都很有意思,普京发出了内心所想

讲者普拉斯
2026-11-20 11:58:25
百亿豪门夫妇正式开撕,发声明互相指责,两人疑未离婚就有新欢

百亿豪门夫妇正式开撕,发声明互相指责,两人疑未离婚就有新欢

开开森森
2026-11-21 09:48:22
今日重要赛事!11月21日,央视CCTV5、CCTV5+节目表

今日重要赛事!11月21日,央视CCTV5、CCTV5+节目表

薇说体育
2026-11-21 10:30:51
字节杀到社交腹地才发现,腾讯的护城河其实没那么深。

字节杀到社交腹地才发现,腾讯的护城河其实没那么深。

放开他让wo来
2026-11-19 22:47:23
深圳街头贴身护墨西哥总统的女警徐瑾,网传年薪百万其实是误会

深圳街头贴身护墨西哥总统的女警徐瑾,网传年薪百万其实是误会

娱乐的硬糖吖
2026-11-20 05:28:23
“游客乱成一团,又冷又饿,接待很一般”,网友发帖称滞留在三峡之巅景区,景区回应:已增派车辆,会把所有游客安全送出景区后再下班

“游客乱成一团,又冷又饿,接待很一般”,网友发帖称滞留在三峡之巅景区,景区回应:已增派车辆,会把所有游客安全送出景区后再下班

南方都市报
2026-11-20 22:42:14
号商下乡收Claude账号:Anthropic算尽了一切,唯独没算到中国号商会上山下乡!

号商下乡收Claude账号:Anthropic算尽了一切,唯独没算到中国号商会上山下乡!

城事堂
2026-11-19 08:49:08
墨西哥总统回程还是搭南航 发文感谢中国女保镖

墨西哥总统回程还是搭南航 发文感谢中国女保镖

看看新闻Knews
2026-11-20 18:56:15
6岁男童被拐17年后,在粥店喝小米粥,突然红眼道:这是奶奶的手艺

6岁男童被拐17年后,在粥店喝小米粥,突然红眼道:这是奶奶的手艺

今天说故事
2025-08-04 18:39:27
14岁从中国赴美,MIT毕业、普林斯顿读博:这位优秀华人女生,为什么被ICE抓走了?

14岁从中国赴美,MIT毕业、普林斯顿读博:这位优秀华人女生,为什么被ICE抓走了?

刚刚讲商业
2026-11-21 08:20:56
深圳 APEC 世纪晚宴菜单曝光,以地道粤菜为主,同时融入国际菜肴元素,菜品设计考量了哪些核心因素?

深圳 APEC 世纪晚宴菜单曝光,以地道粤菜为主,同时融入国际菜肴元素,菜品设计考量了哪些核心因素?

贵重物品爱美食
2026-11-19 22:17:48
WTT球星赛战报!女单4强赛对阵,国乒小将0-3完败,桥本强势零封

WTT球星赛战报!女单4强赛对阵,国乒小将0-3完败,桥本强势零封

南海浪花
2026-11-21 00:33:19
特朗普:美国不支持“台独”!

特朗普:美国不支持“台独”!

陆弃
2026-11-20 11:09:37
从78亿降到32亿,闵行“巨无霸”仍无人举牌

从78亿降到32亿,闵行“巨无霸”仍无人举牌

坠入二次元的海洋
2026-11-21 05:25:52
超级新秀!20岁七号秀!乔丹后第一人!到底有多强?

超级新秀!20岁七号秀!乔丹后第一人!到底有多强?

篮球盛世
2026-11-21 12:12:12
任达华毁容烧声!半边烂脸急就医 狠甩鲜肉男星5巴掌超内疚

任达华毁容烧声!半边烂脸急就医 狠甩鲜肉男星5巴掌超内疚

ETtoday星光云
2026-11-19 10:11:03
惊人预言:中国有一半的人口将塞进这几个地方,房产可能翻番

惊人预言:中国有一半的人口将塞进这几个地方,房产可能翻番

专业聊房君
2026-11-18 19:51:29
2026-11-21 13:31:00
Barrons巴伦
Barrons巴伦
Barron’s《巴伦周刊》官方中文平台,这里有全球资本市场前沿资讯、金融分析、市场洞察和行业解读
1329文章数 34关注度
往期回顾 全部

头条要闻

历时14小时特朗普返回美国 刚下飞机就发布"重大预告"

头条要闻

历时14小时特朗普返回美国 刚下飞机就发布"重大预告"

体育要闻

脚步未停!西班牙摩托车手赤脚徒步入境中国

娱乐要闻

官媒发文 官宣赵丽颖新身份,进步很大

财经要闻

最大回撤超37%,黄金崩盘了吗?

科技要闻

老黄想回来?英伟达董事会:没门

汽车要闻

全新红旗H9最有意思的地方,是它开始有自己的豪华逻辑了

态度原创

本地
旅游
时尚
公开课
军事航空

本地新闻

乌兰察布:以前种土豆,现在种Token

旅游要闻

青岛冬日赏鸥地图|沿着海岸线,赴一场海鸥之约

秋天衣服别总穿纯色,试试这几件条纹单品,简单时尚又减龄

公开课

李玫瑾:为什么性格比能力更重要?

军事要闻

沙特首都第一高楼遭袭:炮弹从天而降

无障碍浏览 进入关怀版