Abusive content notifications: the notifier who gets it wrong incurs liability, even in good faith.

By a judgment of 19 June 2026, the Paris Court holds that whoever obtains removal of content from a platform on the basis of a legal characterisation that will not ultimately be upheld incurs civil liability towards the ousted publisher, regardless of any bad faith. The decision, which extends a line already sketched by the Paris Court of Appeal in 2024, above all offers a method for assessing damages that is likely to be reapplied (TJ Paris, 3rd ch. 2nd sect., 19 June 2026, no. 24/03517).

Removing online content no longer, in most cases, requires going through a judge. A simple notification sent to the platform hosting the disputed publication is enough for that platform, keen to preserve its limited liability regime, to block or remove the targeted content.

This mechanism, designed to enable a swift and effective fight against manifestly unlawful content, has a well-known downside for practitioners: it offers, to those who know how to use it, a formidable instrument of private censorship. A mere allegation of infringement, defamation or trademark violation is enough for a competitor's content to disappear, without adversarial debate or prior judicial review.

The question of the liability of the author of such a notification, when the alleged illegality turns out to be unfounded, had until now been treated with great caution by the courts. The judgment of 19 June 2026 usefully clarifies the applicable regime, laying down a principle whose scope extends beyond the case at hand: the notifier who obtains the removal of content ultimately found to be lawful must repair the harm caused, even in the fullest good faith.

The legal framework for notifying unlawful content has rested for more than twenty years on the French Act of 21 June 2004 on confidence in the digital economy (LCEN), transposing the 2000/31 e-commerce directive. It has been profoundly renovated by Regulation (EU) 2022/2065 of 19 October 2022 on digital services, the DSA, applicable since 17 February 2024. Its Articles 6 and 16 now require hosting service providers to put in place notification and action mechanisms allowing any third party to report content it considers unlawful, with the platform being required to act promptly.

The Paris Court takes care to recall this textual basis, stressing that a request sent to a platform such as YouTube to bring an end to an unlawful publication cannot, in principle, be considered wrongful: to hold otherwise would compromise the very effectiveness of the law by preventing an interested party from seeking to stop an unlawful act from those legally bound to do so.

The notification is therefore, in itself, a perfectly lawful act, and even a desirable one when it targets genuinely unlawful content. The difficulty does not arise from the notification, but from its misuse: invoking an illegality that one knows, or ought to know, to be fragile, solely to silence content that displeases. It is this practice, akin to private justice, that the judgment of 19 June 2026 comes to regulate.

I. The case before the Paris Court.

The author, director and producer of an interview show broadcast on YouTube complained that an association publishing a channel devoted to criticising so-called conspiracy discourse had incorporated, in the title sequence of its own programme, two extracts of two seconds each drawn from two of their shows.

Seeing this as an infringement of their copyright and neighbouring rights, and in the alternative a common-law fault, they had first obtained from the platform the removal of about twenty videos, before suing the association to have the alleged infringement recognised and to obtain a ban on exploiting the disputed shows.

On copyright grounds, the court entirely dismissed the claimants. It recalled that protection presupposes an original work, that is, an intellectual creation of the author reflecting his personality through free and creative choices, and that literary and artistic property does not protect ideas or concepts, but only the original form in which they are expressed.

The elements relied on to establish originality, an orange bench against a black background, the positioning of the participants facing the camera, the alternation of wide and tight shots, amounted, according to the court, to nothing more than an assembly of banal elements drawn from the common stock of television interview techniques. Lacking originality, the format and the shows were not protectable.

The court also rejected the alternative claim of unfair competition and parasitism. Reusing two-second extracts, too brief in themselves to embody any individualised economic value, reflected no intention to ride on the claimants' coattails, but formed part of a legitimate contribution to a debate of general interest. The logo of the original show remained clearly visible, allowing the public to identify the source of the extracts unambiguously.

II. The principle: liability detached from bad faith.

It is on the association's counterclaims that the judgment reveals its full significance. Reasoning on the basis of Article 1240 of the French Civil Code, the court lays down a principle of great clarity: while a takedown request sent to a platform is not wrongful in principle, the blocking of ultimately lawful content causes harm to its publisher, and that harm must be repaired.

The reasoning deserves quotation. The court observes that limiting liability to cases of abuse alone “would leave hanging over everyone the risk of unjustified blockings without any possible recourse, which would not be compatible with freedom of trade, at least where the author is a competitor with an interest in the blocking”. And it continues: “the person on whom the risk of uncertainty must fall can only be the person who creates it”.

The court thus resolutely goes beyond the sole ground of abuse. Article 6 of the LCEN only sanctions, on the criminal side, notifications made in bad faith, that is, those whose author knows to be inaccurate. The court considers that such a criterion, confined to bad faith, would be insufficient to protect the publisher of the blocked content. It substitutes for it, on the basis of common-law civil liability, an objective criterion: the notification is wrongful if, but only if, it seeks a prohibition that is not justified, which must be assessed objectively, according to the legal criteria governing such a prohibition on the merits, and regardless of the good faith of its author.

To support this solution, the court draws on two particularly enlightening analogies. The first is provisional enforcement: whoever enforces a provisional prohibition obtained in court does so at his own risk and must repair the consequences if the prohibition is ultimately held to be unfounded. The court transposes this balance to the situation where the prohibition is obtained without any judicial intervention, through a simple blocking request sent to the technical intermediary.

The second analogy is disparagement: the Commercial Chamber of the Court of Cassation held, on 15 October 2025, that whoever merely reveals to customers an allegation of infringement of a right incurs liability. If a person who merely influences customers is liable, the court reasons, a fortiori is one who directly triggers the cessation of the content's distribution.

The scope of this regime deserves clarification. The court expressly confines the solution to notifications emanating from a person having a commercial interest, and more precisely from a competitor with an interest in the blocking. It is in this case, and only in this case, that the notification is made at the notifier's own risk. The logic is one of allocating the risk of legal uncertainty: as between the publisher of the content, who suffers the blocking, and the competitor who triggers it in order to gain an advantage, it is fair that the latter should bear the burden when the alleged illegality is not established.

III. Assessment of the harm.

In this case, the association claimed 41,200 euros, calculated on the basis of the production cost of the removed videos. The court rejected this claim as a material loss: as the association did not monetise its videos and expected no profit from them, it could not rely on any loss of earnings.

The court did, however, uphold the existence of moral harm, arising from the fact that the blocking deprived the association of the moral benefit it expected from exploiting its videos. Having regard to the number of blocked videos, twenty-two, and to the duration of the blocking, which began in February 2024 and was still in force at the date of the judgment, more than two years later, it fixed this moral harm at 1,000 euros per video, that is, 22,000 euros in total, to be borne by the production company alone, which had originated the blocking requests and alone had an economic interest in them.

The method retained, a lump-sum assessment per blocked piece of content weighted by the duration of unavailability, is easily transposable to other situations. Above all, it hints at much higher amounts in the frequent case of monetised content. Where a channel derives advertising revenue from its videos, the blocking generates a directly quantifiable loss of earnings, which could then be added to the moral harm recognised by the court. A creator or a media outlet whose content generates significant turnover would thus have a considerably larger basis of compensation than the 22,000 euros awarded here.

IV. Practical scope and outlook.

The judgment of 19 June 2026 does not arise in a jurisprudential vacuum. It extends an orientation already visible in a judgment of the Paris Court of Appeal of 5 June 2024, in which a luxury house, having obtained the closure of the Instagram account and the removal of Facebook publications of a watch brand by invoking an infringement ultimately not upheld, had been dismissed from its claims founded on infringement and parasitism and sanctioned on account of its wrongful notification.

The convergence of these two decisions, one delivered in trademark and unfair competition law, the other in copyright law, suggests the emergence of a genuine jurisprudential line: whoever instrumentalises platforms' notification mechanisms to oust a competitor now risks having to repair the resulting harm. This is a major online reputation issue for creators, media outlets and companies whose content is targeted by competitor notifications.

One practical observation must, however, be made regarding the effective implementation of this right to compensation. To act against the author of a notification, one must be able to identify him. Yet it is precisely in notifications founded on infringement of an intellectual property right, copyright, trademarks, designs, that the notifier is forced to reveal himself: claiming a private right presupposes justifying its ownership, and hence disclosing one's identity. Conversely, a large share of the notifications carried out on social media, those invoking a violation of the platform's community guidelines or a hateful or inappropriate character of the content, remain anonymous, which in practice deprives the victim of any recourse against their author.

For litigation practitioners in the digital field, the decision offers a clear roadmap. On the defence side, the ousted publisher now has a solid basis for obtaining compensation when the blocked content turns out to be lawful, without having to demonstrate the notifier's bad faith. On the claim side, companies tempted to use notification mechanisms as a competitive instrument must factor in a new risk: that of incurring civil liability on the sole objective finding that the legal characterisation invoked was unfounded.

In practice, before any notification of a competitor's content, it will now be necessary to assess rigorously the soundness of the legal characterisation retained, to keep a record of the legal reasoning underlying the decision to notify, and to anticipate the potential amount of compensation in case of failure, taking into account the number of contents targeted, the foreseeable duration of the blocking and their possible monetisation.