在11月美国中期选举前夕，社交媒体平台上充斥着由人工智能驱动的虚假信息。由于对言论自由的担忧导致平台削减人工审核团队并转向依赖自动化审核，使得大量虚假内容得以传播 1。前Facebook员工Frances Haugen指出：“现在你可以以极低的成本运行以前不可能达到的质量的信息行动。” 1 Digital Democracy Institute of the Americas（DDIA）的Roberta Braga强调，平台在西班牙语等非英语内容上的审核护栏更少，且随着公司撤回内容审核投资并转向AI，更多虚假信息正“漏网” 1。
得克萨斯州的案例揭示了AI生成谎言如何被快速扩散、变异以及政治对手如何利用这些信息进行攻击 1。民主党参议员候选人James Talarico遭遇AI驱动的虚假信息，称其因心理健康问题住院，后证实为捏造 1。根据DDIA追踪显示，关于Talarico的谎言在西班牙语媒体中演变为“自杀未遂”的说法 1。与此同时，共和党候选人肯·帕克斯顿（Ken Paxton）使用了AI生成的Talarico篡改《圣经》的广告片段 1。
政府机构和主流社交平台在应对此类信息时存在明显的监管漏洞和执行不足 1。在宾夕法尼亚州，国土安全部（DHS）及移民与海关执法局（ICE）副局长Charles Wall公开放大了一名无证移民是民主党候选人Janelle Stelson竞选工作人员的虚假指控 1。尽管Stelson竞选团队证明该移民Chimaobim Adedayo Igbojekwe并未受雇于其团队，但公民责任与伦理华盛顿组织（CREW）指责DHS违反了限制政府雇员政治活动的《哈奇法案》（Hatch Act） 1。目前，Meta要求广告主披露使用合成图像并对AI生成内容进行自动标记；X依靠众包“社区笔记”而非内部安全审查；YouTube允许关于过去选举欺诈的辩论内容存在 1。
Experts warn that artificial intelligence-enabled false narratives are proliferating across social media platforms in the run-up to the November US midterm elections. This trend is exacerbated by tech companies’ shift toward automated moderation and reduced human oversight, driven by concerns over free speech 1. Frances Haugen noted that bad actors can now execute information operations of unprecedented quality at extremely low costs 1.
Specific instances highlight how these technologies are deployed against political candidates. In Texas, Democratic Senate candidate James Talarico was targeted by an AI-generated rumor claiming he had been hospitalized for mental health issues; this falsehood was later proven to be fabricated 1. According to tracking by the Digital Democracy Institute of the Americas (DDIA), the lie regarding Talarico evolved into allegations of a "suicide attempt" within Spanish-language media 1. Republican opponent Ken Paxton further amplified these distortions by using an AI-generated video clip depicting Talarico desecrating a Bible 1. Roberta Braga pointed out that platforms have fewer safety guardrails for non-English content, particularly in Spanish, and that more disinformation is "slipping through" as companies withdraw from investing in content moderation and turn to AI systems 1.
In Pennsylvania, federal agencies were involved in amplifying similar disinformation tactics. Department of Homeland Security (DHS) and ICE Deputy Director Charles Wall publicly amplified a false claim that an undocumented immigrant was working for Democratic candidate Janelle Stelson’s campaign 1. Stelson’s team refuted this by confirming that the individual, Chimaobim Adedayo Igbojekwe, was not employed by the campaign 1. The Citizens for Responsibility and Ethics in Washington (CREW) alleged that DHS violated the Hatch Act, which restricts government employees from participating in political activities 1.
The regulatory frameworks of major tech platforms remain inconsistent and insufficient in addressing these threats. Meta requires advertisers to disclose synthetic images and auto-labels some AI content, whereas X relies on crowdsourced "Community Notes" rather than internal security reviews, and YouTube allows debate content regarding past election fraud to exist 1.
评论
还没有评论，欢迎留下第一条。