胡塞武装密集袭击沙特设施 美军部署AI目标系统协助利雅得防空US Deploys AI Targeting Support to Saudi Arabia as Houthis Intensify Strikes

胡塞武装发言人叶海亚·萨里阿声明，该组织在此轮跨境行动中打击了纳吉兰机场、海米斯穆谢特空军基地、原油泵站等设施，同时在也门南部港口城市亚丁对沙特支持的驻军营地发动了攻势。萨里阿称，行动针对亚丁的纳斯尔营地与一处海军基地，并声称沙特军官正在当地监督军事动员。也门国际承认政府内政部证实，胡塞武装发射的三枚弹道导弹击中了亚丁的两处海岸警卫队设施，其中两枚命中海岸警卫队总部，另一枚击中塔瓦希区的附属建筑。

Houthi spokesperson Yahya Saree stated that the group's coordinated offensive targeted Najran Airport, an airbase in Khamis Mushait, crude oil pumping stations, and military sites in the southern Yemeni port city of Aden. Saree said the fighters struck Al-Nasr camp and a naval installation in Aden where Saudi officers were overseeing military mobilizations. The Interior Ministry of Yemen's internationally recognized government confirmed that three Houthi ballistic missiles hit maritime security installations in Aden, with two missiles striking a coast guard headquarters and a third hitting a coast guard facility in the Tawahi district.

沙特主导的多国联军未立即回应胡塞武装针对具体军事营地的说法。在跨境防空警报发布约10分钟后，沙特民防部门通报威胁已经解除。此外，胡塞武装在该月早些时候曾公开警告国际航空公司避免使用沙特领空，声明仅将麦加和麦地那上空列为豁免区域。胡塞武装控制也门首都萨那及北部与西部大部分领土，该组织近期不仅在也门沿海持续夺取控制区，还通过密集袭击打破了联合国2022年斡旋促成的非正式停火局面。

The Saudi-led coalition did not immediately address the claims regarding the military camps in Aden, while Saudi Civil Defense declared that the direct threat to western cities had passed roughly ten minutes after sounding initial alerts. The rebel group had also warned commercial airlines earlier in the month against entering Saudi airspace, exempting only flight corridors over the holy cities of Mecca and Medina. The Houthis, who control Sanaa alongside expansive sections of northern and western Yemen, have pushed along Yemen's coastline since September while breaking the UN-brokered informal truce reached in 2022.

在沙特王储穆罕默德·本·萨勒曼提出由美军对也门胡塞武装实施直接打击的要求遭美国总统唐纳德·特朗普拒绝后，华盛顿转向提供情报与算法支持。美军派遣100至200名军事人员进驻沙特新组建的联合部队司令部，协助沙特军方操作类似“智能系统”（MSS）的人工智能辅助目标识别平台。该数字系统源自美国防部2017年启动的旗舰级军用人工智能倡议“Project Maven”，此前美军曾在针对伊朗、伊拉克和叙利亚的军事行动中动用该工具实施目标搜索。

Following US President Donald Trump's refusal to launch direct military operations against the Houthis as requested by Saudi Crown Prince Mohammed bin Salman, Washington turned to providing intelligence and software support. The United States deployed 100 to 200 military personnel to assist a new joint force command inside Saudi Arabia, helping Saudi officers operate an artificial intelligence targeting platform similar to the Maven Smart System (MSS). Originating from Project Maven—the US military's flagship artificial intelligence program initiated in 2017—the tool was previously used by American armed forces to identify targets during operations across Iran, Iraq, and Syria.

华盛顿战略与国际研究中心分析指出，该AI平台汇聚、整理并可视化来自实时卫星画面与无人机摄像头的海量情报、监视与侦察数据，自动识别人类分析师耗费数小时难以排查的潜在威胁形态，并标定火箭发射架、坦克或可疑车队，同时测算打击手段。这套被业内比作“战争版谷歌地球”的数字化工具显著压缩了从探测、判别到下达打击指令的“杀伤链”流程。美军今年2月下旬对伊朗实施打击时，系统曾在前24小时内协助锁定了超过1000个目标，远高于此前美军单日约100个目标的处置水平。

Researchers at the Center for Strategic and International Studies explained that the digital platform aggregates, organizes, and visualizes vast streams of intelligence, surveillance, and reconnaissance data from live satellite feeds and drone cameras. The software detects patterns in data that human analysts once struggled to identify manually, flagging potential targets such as rocket launchers, tanks, or suspicious convoys while calculating optimal attack methods. Dubbed by tech commentators as a 'Google Earth for war,' the system dramatically accelerates the kill chain from detection to strike authorization, having assisted US forces in striking more than 1,000 targets within the first 24 hours of operations in Iran compared to a previous baseline of roughly 100 targets per day.

尽管AI目标识别工具能够在数字层面加速情报处理，但其在也门战场应对非对称战术时面临明确的技术限制。胡塞武装在作战中展现出高度的机动性，常态化利用安装在卡车上的车载弹道导弹与小型无人机蜂群发起攻击。该组织作战人员主要隐匿在也门崎岖山乡的丘陵与深邃峡谷中，这种分散机动模式使武装人员往往能在AI系统完成识别并引导打击前迅速撤离现场。

While AI targeting platforms accelerate data processing, users have observed technical constraints when deploying such systems against asymmetric tactics in Yemen. Houthi forces operate with high mobility, deploying truck-mounted ballistic missiles alongside swarms of small unmanned aerial vehicles. The fighters conduct their operations from hills, rugged countryside, and deep canyons, creating tactical windows in which mobile teams can disperse from launch positions before AI-directed reconnaissance completes strike cycles.

此外，算法平台的识别效能在恶劣环境与地面干扰面前会出现衰减。虽然类似MSS的系统在人口稀疏区域表现较佳，但沙漠地形变化、沙尘暴、大风、降雪及茂密植被均会削弱传感器的侦测精度。胡塞武装在实战中布设的各类假目标与欺骗性诱饵，同样会降低AI模型判读的准确率，限制沙特军队将算法优势转化为实际战场压制能力。

Environmental obstacles and deliberate battlefield deception also undermine the reliability of algorithmic detection. Although systems such as MSS are reported to function most effectively across relatively unpopulated expanses, shifts in desert terrain, severe dust storms, high winds, snowfall, and dense vegetation reduce their accuracy. Moreover, deliberately placed decoys on the ground confound automated image analysis, restricting the ability of allied forces to turn computational intelligence into effective battlefield suppression.

在压缩杀伤链周期的同时，AI引导打击系统引发了关于伦理和人道伤亡的严肃关切。尽管相关操作规程要求由人类军事人员对AI识别出的目标进行二次验证并决定处置方式，但在实战高压与高速推进下，技术失误屡有发生。美军此前在伊朗动用类似MSS平台的第一天，系统标定的打击清单中便包含一所学校，导致一枚美军导弹击中伊朗南部的一所小学，造成近200人死亡，遇害者大多数为7至12岁的儿童。类似误击事件表明，算法加速决策机制在实战中潜藏着巨大的平民伤亡风险。

Accelerating the kill chain has raised ethical questions alongside concerns regarding civilian casualties. Although operational protocols stipulate that a human officer must verify targets identified by AI systems before authorizing attacks, practical application has led to devastating errors. When US forces utilized the MSS platform during operations against Iran, the software flagged a school among its targets on the first day. A subsequent American missile strike hit an elementary school in southern Iran, killing nearly 200 people, most of them children between the ages of seven and twelve, illustrating the persistent risks of automated target generation.

除了人道与误击风险外，分析人士指出单纯依靠技术化空袭难以动摇胡塞武装的战略根基。英国皇家国际事务研究所也门问题研究员法雷亚·穆斯利米指出，胡塞武装采取“零和策略”，其创始人主张当代穆斯林社会过于软弱，并将贫困与苦难视作神圣考验。穆斯利米强调，该组织核心意识形态围绕殉道与牺牲展开，美军或沙特即便凭借算法精准摧毁基础设施或导致附带人员死伤，往往不仅无法迫使其屈服，反而容易被胡塞武装借机强化内部动员并巩固对抗意志。

Beyond humanitarian risks, researchers emphasize that technology-driven strikes cannot dismantle the Houthis' operational resolve. Farea al-Muslimi, a Yemen research fellow at the UK think tank Chatham House, explained that the group operates on a 'zero-sum approach' rooted in founding beliefs that contemporary Muslim societies are too soft and that adversity is a divine test. Al-Muslimi observed that because the group's core ideology centers on martyrdom and sacrifice, airstrikes on infrastructure or casualties from collateral damage fail to erode rebel determination and instead reinforce their internal mobilization.