Does Notice and Stay Down Infringe on Fair Use?

The concept of “notice-and-stay-down” has emerged as a proposed evolution of traditional copyright enforcement mechanisms online, shifting from a reactive “notice-and-takedown” model to a proactive obligation on platforms to prevent the reappearance of infringing content once it has been identified. While its advocates in the entertainment, publishing, and software industries frame it as a logical and necessary step to combat persistent infringement, critics warn that such a system risks eroding fundamental protections for lawful speech, particularly the doctrine of fair use. The collision between automated enforcement and nuanced legal exceptions raises deep questions about whether notice-and-stay-down is compatible with the principles of balance embedded in copyright law.

Under the existing notice-and-takedown framework, most prominently codified in the U.S. Digital Millennium Copyright Act (DMCA), a copyright holder who discovers infringing material online can send a takedown notice to the hosting platform. The platform, in turn, removes or disables access to the content to maintain its safe harbor immunity from liability. If the same content reappears, the rights holder must send another notice. Notice-and-stay-down would change this by requiring platforms to implement measures—often automated filtering systems—that detect and block future uploads of the same infringing material, ensuring it does not reappear once removed. The appeal to rights holders is clear: it addresses the so-called “whack-a-mole” problem, where infringing content is repeatedly re-uploaded after each takedown, forcing constant monitoring and repeated enforcement.

The fair use problem arises because copyright infringement is not a strict liability offense; not every use of copyrighted material without permission is illegal. U.S. law recognizes several exceptions, most notably fair use, which allows for limited, transformative uses such as commentary, criticism, parody, news reporting, and educational purposes. A video excerpted for critique, a meme incorporating a movie still, or a mashup that creates new meaning from existing footage may all be legally protected. Determining whether a use is fair is a fact-specific inquiry that requires considering the purpose of the use, the nature of the work, the amount taken, and the effect on the market for the original. This inherently subjective analysis is difficult for humans and virtually impossible for automated systems to perform with any reliability.

In a notice-and-stay-down regime, the risk is that once a rights holder identifies a work as infringing and triggers filtering mechanisms, all future uploads containing that work—or portions of it—will be blocked automatically, regardless of whether those future uses might be fair. For example, a documentary filmmaker attempting to upload a critical analysis of a film that includes short clips for commentary could find their work automatically rejected because the filtering system detects the protected material. Similarly, educators creating online lectures might be unable to share content that incorporates brief excerpts for instructional purposes. The result is a form of prior restraint, where lawful speech is prevented from reaching the public because a machine cannot distinguish between infringing and non-infringing uses.

Advocates of notice-and-stay-down often point to existing content recognition systems, such as YouTube’s Content ID, as evidence that such filtering can work effectively. But even these systems have well-documented issues with overblocking, disputes over false claims, and takedowns of legitimate uses. Content ID and similar tools operate in a legal environment where takedowns are supposed to be contestable and reversible; notice-and-stay-down could harden these processes, making it far more difficult for a legitimate user to challenge an automated block. In effect, the burden shifts from rights holders having to prove infringement to users having to prove non-infringement, often before their work can be published.

This shift has significant implications for the balance of rights in copyright law. The notice-and-takedown system, while imperfect, preserves a measure of responsiveness: content is removed upon notice, but future uses are not automatically suppressed unless and until they too are flagged. Notice-and-stay-down eliminates this second step, embedding the rights holder’s initial claim into an ongoing enforcement mechanism that applies indefinitely. Without robust safeguards, this can create a chilling effect on creativity, discouraging lawful transformative uses for fear of entanglement in automated filters.

The international dimension complicates matters further. In the European Union, Article 17 of the Copyright Directive has already introduced a form of notice-and-stay-down by requiring platforms to make “best efforts” to prevent the availability of infringing works once notified. This has led to the deployment of upload filters across major platforms in the EU. Early evidence suggests that these systems frequently err on the side of overblocking, particularly when dealing with short or transformed excerpts. Because global platforms tend to implement compliance mechanisms in a uniform way, these filtering effects often spill over to users outside the EU, effectively exporting the notice-and-stay-down model to other jurisdictions even where it is not legally mandated.

Proponents argue that the risks to fair use can be mitigated with carefully designed safeguards: transparent claims processes, independent review of disputed blocks, exemptions for certain categories of use, and penalties for abusive or fraudulent claims. Yet history suggests that such safeguards are often underenforced in practice. The sheer scale of user-generated content on major platforms makes individualized review costly and slow, and smaller platforms may lack the resources to implement any meaningful review at all. In such an environment, the path of least resistance is to let the filter’s decision stand, leaving lawful uses stranded behind a digital gate.

The core of the debate is whether the convenience and effectiveness of automated repeat-blocking outweigh the systemic risk to lawful expression. From the perspective of rights holders, notice-and-stay-down is a necessary evolution, closing a loophole that allows infringers to exploit the reactive nature of current enforcement. From the perspective of free expression advocates, it is a step toward privatized, algorithm-driven censorship that treats all uses of copyrighted material as presumptively infringing until proven otherwise.

If notice-and-stay-down is to coexist with fair use, it would require a level of technological nuance and procedural fairness that current enforcement systems have yet to achieve. Until that is demonstrated, the risk remains that in the name of protecting creativity, such a regime could end up constraining it, locking lawful transformative works behind automated filters that cannot see the difference between piracy and parody. The debate is therefore less about whether infringement should be addressed—it should—and more about whether the cure risks silencing the very kinds of expression copyright law was intended to protect.

The concept of “notice-and-stay-down” has emerged as a proposed evolution of traditional copyright enforcement mechanisms online, shifting from a reactive “notice-and-takedown” model to a proactive obligation on platforms to prevent the reappearance of infringing content once it has been identified. While its advocates in the entertainment, publishing, and software industries frame it as a logical…