European Courts Set a Dangerous Precedent for Platform Liability
A German court ruled on September 17, 2026, that Meta is liable under the EU’s Digital Services Act (DSA) for fake advertisements posted by third parties on Facebook and Instagram because of how Meta’s advertising algorithms operate. The ads illegally used another business’s trademarked logo and images of the business’ founder without consent to promote allegedly fraudulent investments. Scam ads are a serious issue for online platforms and their users, but holding platforms liable for third-party content based on how their algorithms rank that content ignores the root of the issue, imposes an unrealistic standard on platforms, and risks disincentivizing the use of algorithms for advertising and other purposes, to the detriment of users and other businesses.
Content-ranking algorithms, which use a variety of data to order content displayed to users, have powered the digital economy for years. These algorithms determine which news content, search results, social media posts, advertisements, and other content users see. In the early days of the Internet, many online services displayed content in chronological order. Today, as content-ranking algorithms have become more sophisticated and accurate, many online services—from social media platforms and search engines to streaming services and advertising platforms—rely on them to provide users with a more engaging online experience tailored to their preferences, habits, and interests.
Because algorithms give online services some control over how third-party content reaches users, governments around the world have reevaluated their online intermediary liability protections, which generally shield online services from legal liability for their users’ behavior, with some limitations. In May 2023, the U.S. Supreme Court heard a case that could have determined whether the United States’ intermediary liability law, Section 230 of the Communications Decency Act, covered algorithmic recommendations. The court ultimately declined to address that question and sent the case back to the lower courts.
Under the DSA, online platforms, including Meta, are not liable for illegal third-party content as long as they had no knowledge of the content and remove the content “expeditiously” upon receiving that knowledge. Platforms may know about illegal third-party content, for example, when the party that posted the content acts under the platform’s authority or control or when the platform receives a notice about the content.
The recent German court ruling goes beyond this, imposing liability on Meta for fake advertisements because the company exercises control over third-party advertising content through its algorithms and advertising practices. The judges based their decision on a European court ruling, in which the court decided that online platforms could lose their liability protections if their algorithms determine which content to show and in what order.
This sets an unworkable standard. By design, algorithms operate automatically; platforms do not manually review and rank every piece of content they display. Treating a platform’s use of algorithms to rank content as evidence that it knows about the underlying content would effectively require platforms to know about all illegal third-party content appearing on their services. That expectation is unrealistic and overly punitive.
Additionally, shifting more of the blame for scam ads onto online platforms fails to address the root of the problem: fraudsters. Punishing platforms does not deter fraudsters from scamming consumers online. Instead, it incentivizes those platforms to fundamentally change the way they operate, to their users’ detriment.
With such a vast amount of content online, algorithms are necessary to efficiently and effectively sort, filter, and recommend relevant content to users. Without content-ranking algorithms, instead of seeing advertisements for products and services tailored to their interests, users would see irrelevant ads displayed to them randomly.
This change would also come at the expense of small or local businesses and nonprofit organizations, which can get more value out of a smaller marketing budget by advertising directly to their target demographic. Meanwhile, platforms themselves will suffer a loss in advertising revenue if they stop using algorithms to target ads to their users. Advertising is a major source of revenue for many free online services, allowing users to benefit from those services without paying a subscription or other fee.
A knowledge standard for online intermediary liability recognizes that online platforms are not aware of every piece of third-party content. Limiting platforms from using a lack-of-knowledge defense for algorithmically ranked content defeats the purpose of the knowledge standard and expands it beyond what’s feasible. It disincentivizes the use of algorithms, does not punish or deter scammers, and decreases the value of online advertising for all parties. In an online world that relies on content-ranking algorithms and targeted advertising, this unreasonable expansion of intermediary liability could undermine important features of the modern Internet.
