根据反虚假信息机构 NewsGuard 本周发布的一份报告，亲俄势力的影响活动导致模仿真实媒体品牌的虚假新闻数量激增了 850%。在受到模仿的 15 家新闻媒体中，有 8 家是法国媒体，其中包括 FRANCE 24。
Pro-Russian influence operations have caused a surge of 850% in fake news stories mimicking real media brands, according to a report published this week by disinformation specialist, NewsGuard. Out of the 15 most imitated news outlets, eight are French – including FRANCE 24. Issued on:
该报告指出：自乌克兰战争爆发以来，俄罗斯的虚假信息传播活动在过去一年中愈演愈烈，这些虚假新闻利用受信任的媒体品牌来破坏公众的信任并欺骗人工智能系统。美国媒体监督机构的专家发现，有三起亲俄势力发起的影响活动制造了 950 篇虚假新闻，这些新闻模仿了 163 个“看似可信的媒体品牌”，且这些虚假新闻在 21 个国家中被传播。
The Russian disinformation operations that emerged at the start of the war in Ukraine have ramped up in the past year, with fake news articles and reports hijacking trusted media brands, to undermine public trust and deceive AI systems, a NewsGuard report released on Monday found. Specialists from the US media watchdog found that three pro-Russian influence operations were the source of 950 fake news stories that mimicked 163 “supposedly trusted news brands” in 21 countries.
在受到模仿的 15 家媒体中，有 8 家是法国媒体；其中 FRANCE 24 在 12 个月内其品牌被虚假新闻利用了 34 次。NewsGuard 的欧洲编辑 Chine Labbé 表示：“法国作为向乌克兰提供援助的最大捐助国之一，成为攻击目标并不令人意外。”她补充说：“每当法国表示对乌克兰的支持（无论是埃马纽埃尔·马克龙的演讲还是对基辅的访问），虚假信息传播活动几乎总是在同一天随之出现。”在 2027 年总统大选前夕，法国似乎更加成为亲俄势力的攻击目标。
Among the 15 most imitated media brands were eight French outlets, including FRANCE 24, which saw its brand coopted by fake news stories 34 times over 12 months. “It's no surprise that France, which is one of the biggest contributors of aid to Ukraine, is being specifically targeted,” said Chine Labbé, Europe editor at NewsGuard. “We know that every time there is a show of diplomatic support for Ukraine – whether it's a speech by Emmanuel Macron, or a trip to Kyiv – disinformation campaigns follow within the same day.”And ahead of the 2027 Presidential elections, France seems evermore in the crosshairs of pro-Russian actors.
三位承诺坚定支持乌克兰的总统候选人——拉斐尔·格卢克斯曼（Raphaël Glucksmann）、加布里埃尔·阿塔尔（Gabriel Attal）和爱德华·菲利普（Édouard Philippe）——最近都成为了与亲俄势力相关的虚假新闻的受害者。
Three presidential candidates who have pledged unwavering support for Ukraine – Raphaël Glucksmann, Gabriel Attal and Édouard Philippe – have all recently been the subject of fake news stories linked to pro-Russian groups.
根据 NewsGuard 的分析，莫斯科发动的信息战具有多重目的。这份报告是由 NewsGuard 与乌克兰的反虚假信息机构、法国政府机构 Viginum 以及匿名研究组织 Antibot4Navalny 共同编制的。
Europe on the front line There are multiple objectives for Moscow’s information war according to NewsGuard, which compiled the report with the Ukrainian agency, Center for Countering Disinformation, French government agency Viginum, and anonymous research group, Antibot4Navalny.
这些行为包括诋毁对乌克兰的支持、破坏西方领导人的声誉、煽动欧洲民主国家内部的紧张局势以及干扰选举进程。
These include discrediting support for Ukraine, undermining the reputation of Western leaders, stoking tensions within European democracies and disrupting electoral processes.
报告发现，模仿知名媒体品牌的假新闻故事同比增加了850%——从2025年上半年的44篇增至2026年上半年的421篇。
Year-on-year, the report found there had been an 850 percent increase in fake news stories mimicking known media brands – rising from 44 in the first half of 2025 to 421 in the first half of 2026.
“这一现象的爆发可能意味着俄罗斯人认为这是侵蚀西方公民对媒体信任的有效策略——或者至少表明，存在以低成本加以利用的巨大潜力，”拉贝表示。
“The explosion of this phenomenon may mean that the Russians see it as an effective tactic for eroding Western citizens' trust in the media – or at the very least, that there is significant potential to be tapped at little cost,” said Labbé.
“如今，利用人工智能创建假新闻网站并生成大量内容几乎不费分文。”报告发现，亲克里姆林宫的影响力行动既模仿了BBC和CNN等全球新闻机构的媒体，也模仿了乌克兰《基辅独立报》和法国《山岳报》等地方新闻网站。
“These days, it costs nothing to create a fake news site using AI and generate a lot of content.”The report found pro-Kremlin influence campaigns had imitated media from global news outlets such as the BBC and CNN, to local news sites such as Ukraine’s Kyiv Independent and France’s La Montagne.
但欧洲媒体机构是这类与俄罗斯政府有关联的混合战争的主要目标，相关行动被称为“套娃”、“Storm-1516”和“分身”。
But European media outlets were the main target for this form of hybrid warfare linked to the Russian government under operations known as Matriochka, Storm-1516 and Doppelganger.
迄今为止被模仿最多的媒体是总部位于法国的多语种平台欧洲新闻台，在2023年3月至2026年9月期间，该台成为82篇假文章的目标。
The most imitated outlet by far was Euronews, a multilingual platform based in France, which was the target for 82 fake articles between March 2023 and September 2026.
2026年的一个例子发表于匈牙利议会选举前不久，那次选举导致亲俄的前总理欧尔班·维克托被赶下台。
An example from 2026 was published shortly before Hungary’s parliamentary elections, which resulted in pro-Russian former prime minister Viktor Orban being removed from power.
该假文章谎称，亲欧洲的候选人、现任总理彼得·马扎尔称美国总统唐纳德·特朗普是“糊涂爷爷”，并称他希望终止两国之间的协议。
It falsely claimed that pro-European candidate and current Prime Minister, Peter Magyar, called US President Donald Trump a “senile grandpa”, and that he wanted to end agreements between their two countries.
在15家被针对最多的媒体中，其余七家法国新闻机构包括《解放报》、《查理周刊》、BFM电视台、法国24台、《费加罗报》、《巴黎人报》和法新社。
The other seven French news outlets among the 15 most targeted include Libération, Charlie Hebdo, BFM, FRANCE 24, Le Figaro, Le Parisien and Agence France-Presse.
“在每一起案例中，俄罗斯的宣传行动都盗用这些品牌的标识和版面设计，有时还使用与真实网站相似的网址创建一个新网站，让普通用户难以辨别其真伪，”NewsGuard写道。“这些报道往往包含捏造的、出自真实专家之口的引语。”
“In every case, the Russian campaigns hijack the brands’ logos and layouts, sometimes creating a new site using a URL similar to the authentic one, making it difficult for the typical user to discern that it is counterfeit,” wrote NewsGuard. “The stories often include invented quotes from actual subject experts.”
“利用一个值得信赖的媒体品牌，可以制造出更可信的假新闻，”拉贝补充道。“但其长期目标并不是让你相信这一条或那一条信息。他们的目的是让我们对所有机构都产生怀疑。”“这可能损害信任，不仅是对我们媒体的信任，而是对所有媒体的信任，”（FMM）编辑战略与合作总监塞西尔·梅吉对NewsGuard表示。该机构旗下包括法国24台和法国国际广播电台。
“Leveraging a trusted media brand makes it possible to produce more credible fake news,” added Labbé. “But the long-term goal isn't to make you believe this or that piece of information. Their aim is to sow doubt in all our institutions.”“It could damage trust, not only in our media, but in all media,” Cécile Mégie, Director of Editorial Strategies and Partnerships at (FMM), which includes FRANCE 24 and RFI, told NewsGuard.
“它正在制造一种对一切都心存怀疑的氛围……挑战在于如何在这种氛围中前行，并表明我们是值得信赖的媒体。”影响人工智能 要在工业规模上传播虚假信息，需要一支由虚假社交媒体账号和亲克里姆林宫媒体组成的“大军”，NewsGuard将其识别为“真理报网络”，也被称为“Portal Kombat”。
“It’s creating a climate of doubt about everything… The challenge is to navigate this climate and to show we are trustworthy media.”Influencing AI Disseminating disinformation on an industrial scale requires an army of fake social media accounts and pro-Kremlin media outlets that NewsGuard identifies as "the Pravda network", also known as Portal Kombat.
该网络由280个网站组成，发布来自亲俄媒体的文章。NewsGuard估计，真理报在2025年发布了约630万篇文章，传播了115条虚假说法。
Made up of 280 websites that publish articles from pro-Russian outlets, Newsguard estimates that Pravda published approximately 6.3 million articles spreading 115 false claims in 2025.
该网络以12种语言发布的一则报道看起来像是来自法国新闻网站《观点》。它在2026年虚假报道称，2万名乌克兰士兵因医疗原因被疏散到法国，随后又逃离军队加入犯罪团伙。
One of the network’s stories published in 12 languages appeared to be from French news site Le Point. It falsely reported in 2026 that 20,000 Ukrainian soldiers had been evacuated to France for medical reasons where they had then deserted the army to join criminal gangs.
官方数据显示，2022年2月至2025年4月期间，略多于200名乌克兰人被转移到法国医院，其中大多数是战伤人员或患有严重疾病的儿童。
Official figures show that a little over 200 Ukrainians were transferred to French hospitals between February 2022 and April 2025, most of whom were war-wounded or children with serious medical conditions.
但铺天盖地的假新闻足以让谷歌的人工智能工具相信这则报道可能是真的。
But the flood of fake news was enough to convince Google’s AI tool that the story could be real.
当被问及此事时，这家科技巨头的人工智能聊天机器人引用了据称来自《Le Point》的虚假文章，声称“数千名”乌克兰士兵已在法国接受了医疗救治。
When questioned, the tech giant’s AI chatbot sourced fake articles supposedly from Le Point to claim that “thousands” of Ukrainian soldiers had received medical care in France.
“这就是‘Matriochka’（该聊天机器人）的目的之一：通过被其他媒体转载或出现在聊天机器人的回复中，来掩盖俄罗斯方面的说法，”拉贝（Labbé）说道。
“That's one of Matriochka's goals: to whitewash the Russian narrative by getting picked up by other media outlets or finding its way into chatbot responses,” said Labbé.
谷歌向NewsGuard辩护称，他们的人工智能工具无法重现那些引用虚假新闻的聊天机器人回复。
Google defended its AI tool to NewsGuard, saying it had been unable to reproduce responses that referenced the fake news story.
“我们对搜索功能中的人工智能功能有着很高的标准，并且在提升回复质量方面投入了大量资源，”谷歌的一位发言人表示。
“We have a high bar for our AI features in Search, and we invest significantly in the quality of responses,” a Google spokesperson said.
这种虚假信息的传播策略旨在“压倒事实核查人员”——他们通过电子邮件向记者发送大量虚假报道，迫使记者去核实这些信息的真实性。
Yet another disinformation tactic aims to “overwhelm fact-checkers” themselves, using email campaigns to bombard journalists with fake reports, urging them to investigate their claims.
根据2025年Vigium的报告，其最终目标是让这些虚假新闻被广泛传播并产生影响力。
The goal – according to a 2025 Vigium report, is that fake stories will get picked up and gain traction.
即使虚假新闻没有在网络上疯传，它们仍然会损害在线新闻的整体质量。拉贝强调：“我们必须谨慎对待某些信息传播方式；对于那些没有引起广泛关注的虚假新闻，我们应该采取不同于对待那些获得数百万浏览量的虚假新闻的应对措施，但我们仍需持续对其进行监控并记录其传播手段。如果我们不加以抵制，人工智能工具可能会受到不良影响，从而导致公众受到错误信息的误导。”
Even if fake stories don’t go viral, they can still erode the overall quality of online news. “We need to be careful about how we address certain narratives,” said Labbé. “We should treat fake news stories that don’t go viral differently from those that generate millions of views, but we still must continue to monitor them and document the tactics being used. If we don’t fight back, there’s a risk that AI tools will be influenced and citizens will be misinformed.”
本文由Joanna York根据法文原文改编。点击此处可阅读原文。
This article was adapted from French by Joanna York. Click here to read the original.