德国一家地方法院裁定,谷歌需为其 AI 生成的搜索摘要中的虚假陈述承担直接责任。
在该案中,谷歌的 AI 错误地将两家出版商与诈骗及不正当商业行为关联起来。
法院将 AI 摘要视为谷歌自身发布的内容,并驳回了谷歌关于用户应自行核实搜索结果真实性的论点。
更新 – 2026 年 6 月 11 日
新增谷歌声明
2026 年 6 月 11 日更新:
谷歌已就这一裁决向我们提供了声明。该公司表示,其 AI 摘要旨在“反映”网络上已有的信息。
“我们大力投入 AI 摘要的质量建设,以确保绝大多数回复都能提供准确信息,并且这些摘要旨在反映网络上已有的信息。我们正在仔细研究这一尚未最终生效的判决,”一位谷歌发言人表示。
谷歌补充说,AI 摘要偶尔会像传统搜索结果一样,遗漏上下文或误解网络内容。但这恰恰是慕尼黑法院裁决的不同之处。法院在 AI 摘要与传统搜索结果之间划清了界限:前者是基于来源松散生成的新内容,而后者则是列出带有直接引用的来源。根据法院的观点,正是这种区别使得谷歌需要承担直接责任。在声明中,谷歌再次重申了法院已驳回的论点,即“用户可以深入挖掘并自行核实”。
原文,2026 年 6 月 9 日:
一家德国法院裁定,谷歌需为其 AI 搜索摘要的内容承担直接责任。此前保护搜索引擎运营商免于承担责任的相关判例法,不适用于 AI 摘要。
慕尼黑地区法院对谷歌发出临时禁令,禁止其通过 AI 生成的搜索摘要散布关于慕尼黑两家出版商的虚假陈述(案件编号:26 O 869/26)。法院将谷歌归类为直接侵权人,因为“AI 摘要”属于其自身发布的内容,而不仅仅是搜索结果列表。
谷歌的 AI 概览曾错误地将两家出版公司与某些搜索查询中的诈骗、订阅陷阱和可疑商业行为联系起来。法院认为,AI 将其他确实可疑的公司的信息与原告混淆,并建立了在任何一个链接来源中都不存在的关联。这两家出版商向谷歌发送了停止侵权函,但谷歌并未做出适当回应。
AI 概览并非搜索结果
法院认为,谷歌的 AI 概览与传统搜索结果的工作方式完全不同。判决书指出,AI 会“用自己的语言,按照自己的结构”重写和评判结果。例如,在本案中,它以“是的,[某公司]以可疑的商业行为而闻名”这样自信的陈述开头,然后构建了自己的结构,包括摘要、针对所谓骗局的警示标志以及给用户的提示。
法院还发现,AI 概览做出了“甚至在搜索结果中都没有出现”的陈述。所有链接来源都没有将原告与 AI 提到的可疑公司建立任何联系。法院称这些是“被告自己的陈述”。
谷歌构建了 AI,谷歌将其提供给用户,因此谷歌对其生成的内容负责,“因为只有谷歌才能影响 AI 的提供方式以及 AI 运行的算法。”
搜索引擎责任规则不适用于 AI“搜索”
法院还审查了德国联邦最高法院的现有裁决,该裁决给予传统搜索引擎和自动补全功能有限的责任豁免。德国联邦最高法院曾认为,搜索引擎运营商仅作为间接侵权者承担责任,因为他们只是让第三方内容变得可被找到。主动检查结果的责任会威胁到搜索引擎的工作方式。
慕尼黑法院认为,这一推理不适用于 AI 概览。常规搜索引擎只是指向外部网站。但 AI 概览通过评估和整合来自各个第三方网站的内容,生成了“独立的、全新的、实质性的陈述”。法院表示,只有谷歌才能检查这些陈述,“至少可以通过将作为基础的第三方网站与其基于这些网站做出的陈述进行比较。”
法院还指出,AI 概览对于使用互联网而言“绝非必不可少”。传统的搜索结果已经能帮助用户筛选信息,AI 概览只是一个额外的功能。
谷歌“用户可自行查证”的辩护理由不成立
在听证会上,谷歌辩称,用户可以自行查看链接来源,以核实 AI 摘要是否正确。该公司声称,用户普遍知道“AI 生成的信息不应盲目相信”。考虑到谷歌提供 AI 概览的规模之大,这一说法颇为引人注目。而且这也不完全属实,因为来源与生成内容之间的联系并非始终存在。
法院驳回了这一观点。通过进一步研究来证伪某个陈述的可能性,并不能“通常免除对该陈述的责任”。AI 概览“本身可以理解”,并且包含“一个独立的陈述,具有独立可理解的内容,且未提及任何其他可能的解释,甚至不可靠的内容”。研究表明,用户几乎从不点击 AI 概览中的来源链接,这支持了法院的推理。
法院将此与新闻法进行了类比,在新闻法中,出版商对本身可理解的摘要(导语)负有责任,即使读者从未阅读过全文。法院指出,如果 AI 概览“普遍被认为不可靠”,谷歌自身的论点也会“显著削弱”该功能的价值。
法院还指出了保护上的漏洞。如果谷歌仅在明显违规时才承担责任,那么当 AI 做出虚假陈述时,受害者将没有真正的法律追索途径。作为来源的第三方网站甚至根本没有做出过相关陈述。因此,受害者无法起诉来源方,而根据现有规则,他们也难以有效起诉谷歌。
因此,谷歌不能援引《数字服务法案》中的托管服务提供商保护条款,也不能退而采用针对搜索引擎的标准“通知-删除”流程。
AI 生成的观点获得的言论自由保护更少
仿佛对谷歌来说其他问题还不够糟糕,法院还针对 AI 生成内容的言论自由保护进行了追究。法院写道,AI 的观点“并非表达者所持信念的体现,而是算法的结果”。
提供 AI 驱动的研究“首先是谷歌商业活动的体现”,并且“至多只是对能够自由表达个人观点和信仰这一兴趣的次要体现”。
在权衡原告的隐私权与谷歌的利益时,谷歌不得不退居次要地位,尤其是因为受到质疑的陈述基于不实的事实。AI 将原告与一些公司关联起来,而根据宣誓证词,这些公司与原告毫无关系。
谷歌承担 80% 的法律费用
法院在大多数指控上支持了原告。法院禁止了关于诈骗、与可疑公司的关联、订阅陷阱、从未发生的电话以及无法联系到等陈述。仅有两项次要请求被驳回。
尽管具体文本已不再显示,但再次违规的风险依然存在。谷歌并未发布带有处罚条款的停止侵权声明,也没有任何机制阻止算法再次生成相同的陈述。谷歌承担 80% 的法律费用;两名原告各承担 10%。
法院认为,该裁决也可能具有国际影响力。
即使 91% 的准确率也意味着数百万个错误答案
慕尼黑的这项裁决意义远不止于这一个案件。AI 初创公司 Oumi 为《纽约时报》进行的一项分析发现,谷歌使用当前 Gemini 3 模型的 AI Overviews 功能,回答正确率为 91%。
这对大多数人日常使用来说已经足够可靠。但以谷歌的规模来看,这仍然意味着每小时会产生数百万个错误答案。如果这些错误内容中有足够多对公司或个人构成诽谤,那么这不仅可能给谷歌,也可能给 ChatGPT、Claude 或 Perplexity 等其他类似服务的提供商带来严重的法律问题。
Oumi 的分析还发现,Gemini 3 给出的正确答案中,有 56% 无法通过谷歌提供的来源得到证实。该人工智能给出的答案,其来源用户无法追溯。
慕尼黑法院正是针对这一问题作出裁决:人工智能会提出自身的主张,而这些主张并不出现在任何链接来源中,运营商必须为此承担责任。这一推理在上诉中能否成立仍有待观察,谷歌也未对该裁决发表评论。但如果该裁决在国际上获得认可,其影响将不仅波及谷歌,还会波及每一个其系统会对网络内容进行改写的 AI 提供商。
A German regional court has ruled that Google is directly liable for false claims in its AI-generated search overviews.
In this case, Google's AI had wrongly linked two publishers to scams and shady business practices.
The court treated the AI overviews as Google's own content and rejected Google's argument that users were responsible for fact-checking the results themselves.
Update – Jun 11, 2026
Added Google Statement
Update, June 11, 2026:
Google has provided us with a statement on the ruling. The company says its AI overviews are designed to "reflect" information that already exists on the web.
"We invest deeply in the quality of AI Overviews to ensure that the overwhelming majority of responses provide accurate information, and they are designed to reflect the information that exists on the web. We’re carefully reviewing this decision, which is not yet final," a Google spokesperson said.
Google adds that AI overviews can occasionally miss context or misinterpret web content, just like traditional search results. But that's exactly where the Munich ruling disagrees. The court draws a line between AI overviews, which generate new content loosely based on sources, and traditional search results, which list sources with direct quotes. That distinction is what makes Google directly liable, according to the court. In its statement, Google also repeats the very argument the court dismissed, that "people can dig deeper and verify."
Original article, June 9, 2026:
A German court has ruled that Google is directly liable for what its AI search overviews say. Previous case law shielding search engine operators from liability doesn't apply to AI overviews.
The Regional Court of Munich hit Google with a temporary injunction barring the company from spreading false claims about two Munich-based publishers through its AI-generated search overviews (case no. 26 O 869/26). The court classified Google as a direct infringer because the "AI overview" is its own content, not just a list of search results.
Google's AI overviews had falsely tied two publishing companies to scams, subscription traps, and shady business practices for certain search queries. According to the court, the AI mixed up information about other, genuinely sketchy companies with the plaintiffs and drew connections that didn't appear in any of the linked sources. The publishers sent Google a cease-and-desist letter, but Google didn't respond appropriately.
AI overviews aren't search results
Google's AI overviews work nothing like traditional search results, the court argues. The AI rewrites and judges results "in its own words and according to its own structure," the ruling says. In the case at hand, for example, it opened with confident claims like "Yes, [company] is known for dubious business practices," then built its own structure with a summary, red flags for the alleged scam, and tips for users.
The court also found that the AI overview made claims "that are not even made in the search results." None of the linked sources drew any connection between the plaintiffs and the shady companies the AI mentioned. The court called these "the defendant's own statements."
Google built the AI, Google offered it to users, so Google owns what it produces, "because it alone has influence over the AI's offering and the algorithms with which the AI operates."
Search engine liability rules don't apply to AI "search"
The court also examined existing rulings from Germany's Federal Court of Justice (BGH), which gave traditional search engines and autocomplete limited liability. The BGH had argued that search engine operators were only liable as indirect infringers because they merely made third-party content findable. A proactive duty to check results would threaten how search engines work.
The Munich court found that this reasoning doesn't apply to AI overviews. A regular search engine just points to outside websites. But AI overviews generate "independent, new, and substantive statements" by evaluating and combining content from various third-party sites. And only Google can check those statements, the court said, "at least by comparing the underlying third-party websites with its own statements based on them."
The court also noted that the AI overview is "by no means absolutely necessary" for using the internet. Traditional search results already help users sort through information, the AI overview is just an extra feature.
Google's "users can check for themselves" defense falls flat
At the hearing, Google argued that users could check the linked sources themselves to verify whether the AI summary was correct. Users generally knew "that information generated with AI should not be blindly trusted," the company claimed. That's a remarkable statement given the scale at which Google serves AI overviews. It's also not entirely true, since the connection between sources and generated content isn't always there.
The court rejected this. The possibility of disproving a statement through further research doesn't "regularly exempt from liability for this statement." The AI overview was "understandable on its own" and contained "a self-contained statement with independently understandable content and no reference to other possible interpretations or even unreliable content." Studies show that users almost never click on sources in AI overviews, which supports the court's reasoning.
The court drew a parallel to press law, where publishers are liable for teasers that are understandable on their own, even if readers never read the full article. Google's own argument would also "significantly diminish" the benefit of the feature, the court noted, if the overview were "generally recognized as unreliable."
The court also pointed to a protection gap. If Google were only liable for obvious violations, victims would have no real legal recourse when the AI makes false claims. The third parties whose websites served as sources hadn't even made the statements in question. So victims couldn't sue the sources, and under existing rules they couldn't effectively sue Google either.
As a result, Google couldn't invoke host provider protections under the Digital Services Act or fall back on the standard notice-and-take-down process for search engines.
AI-generated opinions get less free speech protection
As if the rest wasn't bad enough for Google, the court also went after free speech protection for AI-generated content. An AI's opinion is "not the expression of an acquired conviction of the persons expressing it, but the result of an algorithm," the court wrote.
Offering AI-powered research is "above all an expression of Google's business activities" and "at most a secondary expression of an interest in being able to freely express one's opinion and beliefs."
When weighing the plaintiffs' privacy rights against Google's interests, Google had to take a back seat, especially since the challenged statements were based on untrue facts. The AI had linked the plaintiffs to companies that, according to sworn affidavits, had no connection to them whatsoever.
Google picks up 80 percent of the legal tab
The court ruled in favor of the plaintiffs on most counts. It banned claims about scams, connections to dubious companies, subscription traps, phone calls that never happened, and lack of availability. Only two minor requests got denied.
The risk of repeated violations remained, even though the specific texts were no longer being displayed. Google hadn't issued a cease-and-desist declaration with a penalty clause, and nothing stopped the algorithms from generating the same statements again. Google covers 80 percent of the legal costs; the plaintiffs pay 10 percent each.
The ruling may also have international reach, according to the court.
Even a 91 percent accuracy rate means millions of wrong answers
The Munich ruling goes far beyond this one case. An analysis by AI startup Oumi for the New York Times found that Google's AI Overviews with the current Gemini 3 model answered correctly 91 percent of the time.
That's solid enough for everyday use by most people. But at Google's scale, it still means millions of wrong answers every hour. If enough of that wrong content defames companies or individuals, it could become a serious legal problem not just for Google but for other providers of similar services like ChatGPT, Claude, or Perplexity.
The Oumi analysis also found that 56 percent of the correct Gemini 3 answers couldn't be backed up by the sources Google linked. The AI is giving answers whose origins users can't trace.
The Munich court tackled exactly this problem: the AI makes its own claims that don't appear in any linked source, and the operator has to answer for them. Whether this reasoning holds up on appeal remains to be seen, and Google hasn't commented on the ruling. But if it gains traction internationally, the fallout could hit not just Google but every AI provider whose systems paraphrase content from the web.