在针对CEO的袭击事件、一场以数据中心为目标的全国性抗议运动,以及人们对AI取代工作岗位的担忧日益加剧之后,联邦情报机构和国内执法部门正在传阅报告,将一个新的国内目标纳入视野:反科技极端分子。
WIRED获取的来自国土安全部、联邦调查局及联合反恐情报中心(fusion centers)的超过1000页未公开报告显示,一场全国性的转变正在发生,旨在监控这一被认定为新兴威胁的、令人担忧且范围宽泛的人群及其活动。
这项新举措紧随唐纳德·特朗普总统的《国家安全总统备忘录第7号》之后,该备忘录指示司法部针对任何持有“反美”、“反基督教”和“反资本主义”信仰的人。本月早些时候,特朗普的反恐事务主管塞巴斯蒂安·戈尔卡发布了一项公开反恐战略,声称左翼极端分子是美国面临的三大反恐优先事项之一。
综合来看,特朗普政府的这些指令已征用国内监控机制,用以监视那些挑战白宫意识形态的言论和集会,并将其定罪。对反科技极端主义的新关注,为这个在政治和物质资本上大力投资AI及数据中心建设的总统任期下已有的公开认定类别,增加了一个此前未被报道的范畴。
在WIRED获取的这批文件中,有一份纽约情报与反恐局的报告,该报告警告称,AI的采用将引发广泛的动荡。尤其值得注意的是,该局为一种据称是新兴的极端主义威胁提出了一个新术语。
报告写道:“未来五年内,新兴AI技术可能导致的混乱氛围,或将助长大规模抗议活动,进而演变为内乱和反科技暴力极端主义活动,尤其是在纽约市这样的大型城市区域。”“反科技暴力极端主义”这一术语并未出现在任何公开的国土安全部或联邦调查局的国内极端主义报告或指南中,它代表了一种将多种意识形态归入单一极端主义类别的新分类方式。
在同一份情报局的评估报告中,分析人员还描述了一种在极端理性主义者齐兹·劳塔被捕和受审后出现的新型威胁。此人据称领导着一个类似邪教的小团体,该团体中有三名成员因与一种痴迷于人工智能存在风险的意识形态有关而被指控犯有谋杀罪。
尽管齐兹派的意识形态本质上是极端主义的,但围绕人工智能可能带来灾难性后果的同样恐惧,其不那么极端的版本却是人工智能对齐专家、机器学习工程师乃至前沿人工智能公司普遍关注的问题。尽管如此,情报局警告称,由于齐兹派成员“试图论证一种信念,即神一般的人工智能化身即将降临”,并认为“人类必须在当下充分利用时间,致力于确保其符合人类道德,否则将因未能做到这一点而面临存在性后果”,因此,在齐兹派成员受审后,“关于人工智能的偏执观点”可能会蔓延开来。
纽约警察局的情报评估是在该部门去年与联邦调查局合作,监控一个活动团体协调志愿者在纽约移民法庭旁听公开听证会的 Signal 聊天记录之后进行的。根据《卫报》获得的文件,联邦调查局监控这些活动人士,是作为针对“无政府主义暴力极端分子”更广泛调查的一部分,而“无政府主义暴力极端分子”正是新反恐战略中列出的威胁类别之一。
自“9·11”事件后建立以来,全美现已遍布 80 个情报融合中心,它们充当着联邦情报机构与州及地方执法部门之间的中间人。除了对部分美国民众因人工智能快速普及而感到不安的担忧外,这些中心还在收集和传播有关据称针对数据中心的威胁的“情报”。
例如,宾夕法尼亚州西部的一个联合情报中心声称,“包括国家支持的行为体、犯罪集团以及极端分子(如本土暴力极端分子或环境极端分子)在内的敌对行为体,可能会以美国数据中心为目标”,并且“这些行为体还可能利用数据中心对美国经济所具有的战略重要性,将其用于加密货币挖矿等活动,或利用空壳公司等第三方实体来获取美国数据和基础设施的访问权限。”
北弗吉尼亚地区情报中心的一份报告警告称,受政府相关不满情绪和阴谋论影响的“反政府、反权威暴力极端分子”(AGAAVEs),已针对数据中心及其他关键基础设施设施展开行动前策划,意图破坏政府运作。但法律专家表示,在可疑活动报告指标的分类中,该情报报告所列出的活动很容易由和平抗议者实施。
“这些情报报告是机构长期以来将抗议活动,甚至仅仅是持有强烈意见,视为暴力前兆这一传统的一部分,”美国全国有色人种协进会法律辩护基金高级法律顾问斯宾塞·雷诺兹告诉《连线》杂志。“可疑活动报告极其不可靠,往往针对模糊或无辜的行为,且是在宽松的标准下发布的。这些报告通常数量庞大,使得执法人员能够注入自身偏见,在事实中看到他们想看到的东西。”
北弗吉尼亚情报中心标记为可疑的、定义模糊的活动包括:“表达/暗示的威胁”、“观察/监视”、“拍照”、“测试/试探安保措施”以及“试图入侵”。
“联邦调查局调查那些实施或意图实施暴力及犯罪活动,且该活动构成联邦犯罪或对国家安全构成威胁的个人,”联邦调查局在给《连线》杂志的一份声明中写道。“我们没有进一步的评论。”
国土安全部未回应《连线》杂志的置评请求。
与此同时,同一情报中心在三月份还传阅了一份报告,显示其监控了与对科技持批评观点相关的受宪法保护的活动和示威。这些活动包括多场针对埃隆·马斯克洗劫美国政府的“特斯拉下架”抗议活动,以及一场由“驱逐埃尔比特”组织赞助的“与科技狂怒决裂”活动,该组织是一个旨在阻止对以色列武器制造商埃尔比特进行投资的激进团体。
除了在融合中心和联邦执法机构内部工作的情报分析人员外,与联邦执法机构签约的开源情报公司似乎也在网络上搜寻他们所谓的反科技情绪。2025年1月,SITE 情报公司向融合中心发布简报,声称一个“新勒德主义”Discord 服务器中的对话已转向暴力,该群组的一名用户呼吁对科技公司 CEO 和发电厂实施暴力。
“SITE 是一家营利性私人情报公司,为其执法客户监控社交媒体。它承诺完成一项极其困难、甚至不可能完成的任务,即持续挖掘由匿名发帖者撰写的、充满内部笑话、俚语、不同语言、含糊其辞等内容的社交媒体信息,以提供能够预测威胁的可靠情报,”雷诺兹说。“相反,这类活动往往侧重于人们对警务、堕胎、经济不平等、疫苗或任何其他当今热点话题的看法。”
“通过将我们的开源情报工作范围严格限定在与现实世界危害有确凿联系的社群上,即使是挑衅性的言论也具有信息价值,能够展示社群内部对某个目标的态度。我们的报告显示,网络上鼓吹破坏数据中心的威胁言论显著增加,这确实值得担忧,”SITE 创始人丽塔·卡茨在一封电子邮件中告诉《连线》杂志。
WIRED 获得的文件还显示,融合中心目前正在监控线下集会。北弗吉尼亚融合中心就当地公民活动中的示威活动生成了一份报告,其中包括阿灵顿县预算会议和费尔法克斯县学校董事会会议。在全国范围内,市政厅会议和预算委员会会议已成为当地居民表达对社区内建设数据中心不满的主要场所。
根据 AI 安全公司 10a Labs 追踪反对数据中心建设的项目“数据中心观察”(Data Center Watch),42 个州的数百个组织已组织起来,阻止在其城镇和县内建设数据中心。这些努力往往充满争议。在加利福尼亚州、伊利诺伊州、印第安纳州、新泽西州、俄克拉荷马州和威斯康星州,州及地方警察已将市政厅会议上批评数据中心的发言者驱逐或逮捕,在其中一个案例中,发言者甚至还未被允许发言。
根据美国法律,国内恐怖主义并非在审判中适用的独立罪名。相反,国内恐怖主义法律允许对极端分子进行定位和监控,相关指控有时会附带恐怖主义加重量刑,有时则完全不包含。这导致抗议者和活动人士在受到国内极端主义条款监控的同时,被以非法侵入和故意破坏财产等罪名起诉。
联邦机构对反科技活动的关注,从一份在全国融合中心流传的极端主义研究员毛罗·卢布拉诺(Mauro Lubrano)讲座邀请函中可见一斑。卢布拉诺已成为反科技极端主义领域最顶尖的专家之一。他是《停止机器:反科技的崛起》(Stop the Machines: The Rise of Anti-Technology)一书的作者,该书描述了一种新近提出的威胁矩阵中的三大主要分支:叛乱无政府主义者、生态极端主义者和生态法西斯主义者。
卢布拉诺的著作将“大学炸弹客”泰德·卡钦斯基的追随者、德国无政府主义者、墨西哥生态极端分子以及“恐怖电报”集体中的极右翼法西斯分子,视为新兴技术极端主义运动中虽各有不同但相互关联的组成部分。在卢布拉诺的分析中,这些团体的共同点在于,他们都曾策划或实施过旨在推进其意识形态目标的暴力行为。
卢布拉诺表示,他对自己的讲座内容出现在融合中心并不感到惊讶,但他也警告说,任何反技术极端主义的框架都必须谨慎运用。“我希望我和其他同事提出的警告能够得到重视。虽然反技术暴力是不可接受的,但这不应被用作借口,将人工智能和新兴技术安全化,从而压制那些对当前发展轨迹持批评态度的人,”卢布拉诺告诉《连线》杂志。
但斯宾塞·雷诺兹表示,尽管这些团体构成了真实(尽管有限)的威胁,但像“反技术极端主义”这样的类别可能被定义得过于宽泛,以至于将和平的数据中心抗议者、人工智能怀疑论者,以及任何对渗透现代生活的技术有意见的人都囊括其中。
“随着人们继续为更美好的未来而组织起来,我们很可能会看到对这种反对声音进行更多的监控和定罪,就像近几十年来我们对‘黑人的命也是命’、‘占领华尔街’和环保运动所做的那样,”雷诺兹说。
美国国土安全部情报与分析办公室2025年1月的一份报告进一步强化了这一观点,试图将联合健康保险公司首席执行官布莱恩·汤普森的疑似刺杀者路易吉·曼吉奥内与卡钦斯基联系起来。“执法部门报告称,此人可能从泰德·卡钦斯基(大学炸弹客)及其反技术信仰中汲取了灵感,”报告写道,但没有提供进一步证据。报告最后发出警告,声称当高管们“被认为是在利用弱势群体”时,他们“面临针对性暴力行为或暴力威胁的风险更高”。
但或许最能体现非暴力技术批评如何被卷入并标记为威胁的案例,出现在 SITE Intelligence 于 2025 年 4 月流传的一份开源报告中。该报告将进步派非营利组织“更完美联盟”发布的一段视频标记为威胁,该视频讲述了佐治亚州一座数据中心对附近居民造成的破坏性影响。视频中没有任何内容鼓吹针对财产或人员的暴力行为。但得益于融合中心的定向监控,这个倡导组织如今已作为潜在威胁源,在全美情报与执法机构间流传。
本文最初发表于 WIRED.com。
In the wake of attacks on CEOs, a nationwide protest movement targeting data centers, and increasing concerns about AI job replacement, federal intelligence agencies and domestic law enforcement are circulating reports with a new domestic target in mind: anti-technology extremists.
More than 1,000 pages of unpublished reports from the Department of Homeland Security, FBI, and fusion centers obtained by WIRED show a national shift taking place to surveil this new and worryingly broad category of people and activities deemed an emerging threat.
This new effort follows President Donald Trump’s National Security Presidential Memo 7, which instructs the Department of Justice to target anyone holding “anti-American,” “anti-Christian,” and “anti-capitalism” beliefs. Earlier this month, Trump’s counterterrorism czar, Sebastian Gorka, released a public counterterrorism strategy claiming that left-wing extremists are one of the three top counterterrorism priorities facing the United States.
Taken together, these Trump administration directives have commandeered the domestic surveillance apparatus to surveil and criminalize speech and assembly that challenges the ideology of the White House. A new focus on anti-technology extremism adds an unreported category to already public designations under a presidency that has heavily invested political and material capital in AI and data center proliferation.
Among the documents in the tranche obtained by WIRED is a New York Intelligence and Counterterrorism Bureau report that warns of widespread upheaval in response to AI adoption. Of particular note is a novel term for what the bureau purports to be an emerging extremism threat.
“The chaotic atmosphere that may result from emergent AI technology in the next five years may fuel large-scale protests that devolve into civil unrest and anti-tech violent extremist activity, especially in large urban areas such as New York City,” the report reads. The term “anti-tech violent extremism” does not appear in any publicly available DHS or FBI domestic extremism reports or guides and represents a novel grouping of a wide range of ideologies under a single extremist category.
In the same Intelligence Bureau assessment, analysts also describe a novel threat emerging in the wake of the arrest and trial of Ziz Laota, an extreme rationalist who allegedly led a small cultlike group, three members of which have been charged with murder, tied to an obsessive ideology focused on the existential risk posed by AI.
While the Zizian ideology is extremist in nature, a less extreme version of the same fears surrounding the cataclysmic potential of AI are a common concern among AI alignment experts, machine-learning engineers, and even frontier AI companies. Nonetheless, the Intelligence Bureau warns that “paranoid views regarding AI” may proliferate in the aftermath of the Zizians’ trial, thanks to their “attempt to reason the belief that a godlike incarnation of AI is imminent,” and belief that “humans must best use their time in the present to devote themselves to ensuring its compliance with human morality, or face existential consequences for failing to do so.”
The NYPD intel assessment follows the department’s collaboration with the FBI last year to monitor the Signal chat of an activist group coordinating volunteers to monitor public hearings at immigration courts in New York. According to documents obtained by The Guardian, the FBI surveilled activists as part of a broader investigation into “anarchist violent extremist actors,” one of the threat categories named in the new counter terrorism strategy.
Created in the wake of 9/11, 80 fusion centers now pockmark the country and serve as go-betweens for federal intelligence agencies and state and local law enforcement. In addition to concerns about portions of the American populace disturbed by the rapid proliferation of AI, these centers are also gathering and circulating “intelligence” about alleged threats to data centers.
A Western Pennsylvania fusion center, for example, claimed that “adversarial actors, including state-sponsored entities, criminal groups, and extremists, such as homegrown violent extremists or environmental extremists, may target US data centers” and that “these actors could also exploit the strategic importance of data centers to the US economy, using them for activities like cryptocurrency mining or leveraging third-party entities, such as front companies, to gain access to US data and infrastructure.”
A report from the Northern Virginia Regional Intelligence Center warned that AGAAVEs—anti-government, anti-authority violent extremists—influenced by government-related grievances and conspiracy theories, have engaged in pre-operational planning targeting data centers and other critical infrastructure facilities to disrupt government operations. But in the breakdown of Suspicious Activity Reporting indicators, the intelligence report lists activities that could easily be carried out by peaceful protesters, legal experts say.
“These intelligence reports are part of a long tradition of agencies identifying protest or even simply having strong opinions as precursors to violence,” Spencer Reynolds, senior counsel at the NAACP Legal Defense Fund, tells WIRED. “Suspicious activity reports are incredibly unreliable, often about vague or innocent behavior, issued under permissive standards. These reports, often received in large volumes, allow officers to inject their own biases and see what they want to see in the facts.”
Among the vaguely defined activities flagged by the Northern Virginia intelligence center as suspicious are “expressed/implied threat,” “observation/surveillance,” “photography,” “testing/probing of security,” and “attempted intrusion.”
“The FBI investigates individuals who commit or intend to commit violence and criminal activity that constitutes a federal crime or poses a threat to national security,” the FBI wrote in a statement to WIRED. “We have no additional comment.”
The Department of Homeland Security did not respond to WIRED’s request for comment.
Meanwhile, the same intelligence center also circulated a report in March showing monitoring of constitutionally protected events and demonstrations related to critical views on technology. These events included multiple “Tesla Takedown” protests against Elon Musk’s ransacking of the US government and a “Break Up With Tech Rager” sponsored by Eject Elbit, an activist group organizing to halt investment in the Israeli weapons manufacturer Elbit.
In addition to intelligence analysts working inside fusion centers and federal law enforcement agencies, open source intelligence companies that contract with federal law enforcement agencies appear to be scouring the web for what they claim to be anti-technology sentiment as well. In January 2025, SITE Intelligence circulated bulletins to fusion centers alleging that conversations in a “neo-Luddite” Discord server had turned violent, with one user of the group calling for violence against tech CEOs and power plants.
“SITE is a for-profit private intelligence firm that monitors social media for its law enforcement customers. It promises to do an incredibly difficult if not impossible job, consistently mining social media written by anonymous posters, full of in-jokes, slang, different languages, vagueness, and so on, to deliver credible information that can predict threats,” Reynolds says. “Instead, this type of activity tends to focus on people’s views about things like policing, abortion, economic inequality, vaccines, or any other hot-button topic of the day.”
“By narrowing our OSINT focus exclusively to communities with a proven link to real-world harm, even trolling remarks have an informative value, demonstrating sentiment within a community toward a target, and our reports have shown a notable spike in online threats advocating for sabotage against data centers, which is a true cause for concern,” Rita Katz, founder of SITE, tells WIRED in an email.
The documents obtained by WIRED also show that fusion centers are currently keeping tabs on in-person assemblies. The Northern Virginia center generated a report about demonstrations at local civic events, including the Arlington County budget meeting and the Fairfax County School Board meeting. Across the country, town halls and budget committee meetings have been among the chief forums for local residents to express their dissent with data centers being built in their neighborhoods.
According to Data Center Watch, a project by AI security firm 10a Labs that tracks opposition to data centers, hundreds of organizations across 42 states have organized to block data center construction in their towns and counties. These efforts are often contentious. In California, Illinois, Indiana, New Jersey, Oklahoma, and Wisconsin, state and local police have removed or arrested speakers at town halls who criticize data centers, in one case before they were even allowed to speak.
Under US law, domestic terrorism is not a stand-alone crime that is brought to bear during trial. Instead, domestic terrorism laws allow for targeting and surveillance of extremists, with charges sometimes bearing terrorism enhancements and sometimes excluding them altogether. This has led to protesters and activists being surveilled under domestic extremism provisions while being charged with crimes like criminal trespass and vandalism.
The zeroing in on anti-tech activity by federal agencies is evident in an invitation to a lecture by extremism researcher Mauro Lubrano circulating in fusion centers across the country. Lubrano has emerged as one of the foremost experts on anti-technology extremism. He is the author of Stop the Machines: The Rise of Anti-Technology, which describes three main strains of a newly minted threat matrix: insurrectionary anarchists, eco-extremists, and ecofascists.
Lubrano’s book identifies followers of “Unabomber” Ted Kaczynski, German anarchists, Mexican eco-extremists, and far-right fascists in the Terrorgram Collective as distinct but aligned components of an emerging tech extremism movement. In Lubrano’s analysis, these groups are united by the fact that they have all plotted or carried out acts of violence in furtherance of their ideological goals.
Lubrano said he was not surprised that his lecture turned up in a fusion center but cautioned that any anti-tech extremism framework must be exercised carefully. “I hope the warning I, along with other colleagues, raised is being acknowledged. While anti-technology violence is unacceptable, it should not be used as an excuse to securitize AI and emerging technologies, thereby silencing those who are critical of the current trajectory,” Lubrano tells WIRED.
But Spencer Reynolds says that, despite the real, if limited, threat posed by these groups, a category like “anti-tech extremism” could be drawn so broad as to ensnare peaceful data center protesters, AI skeptics, and anyone with a bone to pick with technology that permeates modern life.
“As people continue to organize for a better future, we’re likely to see more surveillance and criminalization of this opposition, just as we have of Black Lives Matter, Occupy Wall Street, and environmental movements in recent decades,” Reynolds says.
A January 2025 DHS Office of Intelligence and Analysis report furthers this perspective by attempting to connect Luigi Mangione—the alleged assassin of UnitedHealth CEO Brian Thompson—with Kaczynski. “Law enforcement reports that the individual may have drawn inspiration from Ted Kaczynski (the Unabomber) and his anti-technology beliefs,” the report reads, without offering further evidence. It concludes with a warning alleging that executives “are at a heightened risk for targeted acts of violence or threats of violence” when they are “perceived as taking advantage of individuals of lesser means.”
But perhaps the clearest-cut example of how nonviolent critiques of technology can be swept up and flagged as a threat is found in an open source report circulated by SITE Intelligence in April 2025. The report flags a video from the progressive nonprofit More Perfect Union on the destructive effects of a data center to nearby residents in Georgia. Nothing in the video advocated for violence against property or people. But thanks to fusion center targeting, the advocacy group is now circulating among US intelligence and law enforcement across the country as a potential threat vector.
This story originally appeared at WIRED.com.