Germany’s NetzDG and the Threat to Online Free Speech

Who are the arbiters of free speech? As voices clamor to be heard online, the answer is, increasingly, the giants of social media. Facebook, Twitter, and other social media platforms have become the engines of online speech and the primary mode by which we access news and information. As online users express themselves on private platforms but within the public sphere, governments are increasingly calling upon social media companies to censor speech that violates domestic law.

Germany’s Network Enforcement Law, or NetzDG,[1] underscores the risks of this approach. The NetzDG was passed in June 2017 to combat terrorist and extremist content online and recently went into effect.  It requires social media companies to block or remove content that violates one of twenty restrictions on hate and defamatory speech in the German Criminal Code.[2] Registered social media users may submit complaints through procedures established by social networks in accordance with the NetzDG.[3] Social networks must remove “manifestly unlawful” content within twenty-four hours of receiving a complaint, and all other unlawful content within seven days, with limited exceptions.[4] The NetzDG applies to all “profit-making internet platforms that are intended to allow users to share content with other users or make it publicly available,” and largely relies on companies to enforce the law.[5] Companies that repeatedly fail to comply with the NetzDG may be fined up to fifty million Euros.[6]

In effect, the NetzDG conscripts social media companies into governmental service as content regulators.[7] Social media platforms not only must monitor and review content, but also must interpret the German Code’s byzantine and sometimes ambiguous provisions. But unlike a true government agency, social media companies face steep penalties for under-enforcement. As a result, the NetzDG incentivizes intermediaries to overpolice speech—social media companies are more likely to remove demeaning content that could potentially violate the Criminal Code than risk a fifty-million-Euro fine;[8] indeed, many of these transnational companies may even be duty-bound to U.S. shareholders to construe the speech categories broadly to avoid liability.

Overpolicing driven by risk aversion is further exacerbated by the sheer number of user-generated posts.  The requirement that social media companies remove content within 24 hours encourages social networks to immediately remove content that appears to fall into proscribed categories of speech, leaving companies little time to consider questionable content.[9]   

Beyond incentivizing overenforcement, the NetzDG further threatens online free speech due, in part, to the breadth of Germany’s defamation law. German defamation law encompasses a much broader set of categories than the U.S. corollary. In accordance with the German Criminal Code, the NetzDG requires social media platforms to remove two categories of speech: defamation, which roughly equates to communications that injure reputation, and “insults,” which roughly equates to communications that harm an individual’s personal honor. While Germany’s defamation law creates liability for false statements of fact that cause reputational harm, the German law of insult enables plaintiffs to sue for true statements that communicate an offender’s “own lack of respect for the victim.”[10]  In contrast, U.S. defamation law prohibits only false comments that could “lower [an individual] in the estimation of the community or . . . deter third persons from associating or dealing with him.”[11] Accordingly, the German law of insult encompasses a significant amount of critical or demeaning speech that would be protected under the First Amendment in the United States.

For instance, the German equivalent of “the middle finger”—i.e. the “tapping of the index finger on the forehead” (called “the bird”)—is widely understood to be an illegal insult in Germany.[12] A video or image of “the bird” posted to another’s Facebook profile could conceivably be actionable under the NetzDG. As another example, “[i]t can be a criminal offense in Germany to call another person a ‘jerk,’ or even to use the informal du, or ‘thou,’” to communicate a lack of respect for the recipient.[13] German courts have penalized other similar “derogatory epithets,” including “asshole,” “bastard,” “old goat, “fathead,” “dope,” and “whore.”[14]

Contrast these examples with defamatory comments in the United States. To establish defamation, “it is never sufficient to prove merely that the statements made about [the] plaintiff are annoying or embarrassing.”[15] “For example, it is not defamatory to publish a photograph [a] plaintiff deems unflattering,”[16] to “[depict] a former employee as financially indebted to the terminating employer,”[17] or to “portray someone as somewhat callous, insensitive, and unsympathetic.”[18] And as the First Circuit explained, “hyperbole and expressions of opinion unprovable as false are constitutionally protected.”[19] More importantly, U.S. defamation law does not cognize a right against insult. A personal insult in the confines of one’s home—or in a private Facebook message—does not injure reputation.    

The increased chilling effect of Germany’s prohibition on insults bears out in practice. Germans bring far more complaints of insult than defamation or slander. In 2013, German courts adjudicated just 460 criminal defamation cases and 417 criminal slander cases, about half of which resulted in fines.[20] By contrast, German courts heard 26,757 criminal insult cases that year, leading to 21,454 convictions and 20,390 fines.[21] This discrepancy may stem from the ease of filing an insult complaint.[22] And, while the German legal profession has attempted to “limit the actual reach of the law of insult,” insult retains a “far richer life in the popular mind than in the courts,” increasing the likelihood that social networks will receive insult complaints under the NetzDG.[23]

Although German insult claims often ultimately fail,[24] their mere existence present risk of liability for social networks that fail to remove potentially unlawful speech.[25] Social media platforms will need to carefully evaluate content to determine if the speaker intentionally communicated a lack of respect for the victim—even in the absence of demonstrable reputational harm. In close cases, social media companies will likely err on the side of caution in order to avoid penalties under the NetzDG.

The NetzDG underscores the costs of imposing intermediary liability for unlawful content online. By incorporating Germany’s defamatory and broader insult provisions, the law encourages social media platforms to overpolice speech—chilling legitimate individual expression and discourse on today’s most significant forums for public debate.

[1] For a German-language version of the statute, see Beschlussempfehlung und Bericht [Resolution and Report], Deutscher Bundestag: Drucksache [BT] 18/13013, (Ger.) [hereinafter NetzDG].

[2] See Strafgesetzbuch[StGB] [Penal Code], translation at (Ger.).

[3] NetzDG, supra note 1, § 3(1)-(2).

[4] Id. § 3(2).

[5] Id. § 1(2).

[6] Id. § 4(1)-(2).

[7] While the NetzDG enables social networks to transfer complaints to government-approved institutions for review, the law authorizes the German Federal Office of Justice to revoke an institution’s approval or prohibit a social network from referring complaints to a particular body. See id. §§ 3(5)-3(9). Moreover, although German authorities must obtain a court order before levying a fine for unlawful content, they may avoid this provision by imposing fines for violations of the law’s procedural requirements, which do not require judicial oversight. See id. § 4(5).

[8] Facebook already employs a team of censors “to distinguish between hate speech and legitimate political expression,” occasionally leading to paradoxical results. E.g., Julia Angwin & Hannes Grassegger, Facebook’s Secret Censrship Rules Protect White Men from Hate Speech but Not Black Children, ProPublica (June 28, 2017, 5:00 AM), The NetzDG would impose additional and more stringent requirements on social media companies to remove broader categories of what might constitute hate speech or defamation under German law. 

[9] Facebook has stated that it currently “remove[s] more than 80% of the reported [hate speech] which has been classified as illegal by German non-governmental organizations . . . .” As it has explained, “In the past two months, we erased an average of 3.500 posts per week, [sic] which were reported as hate [speech] in Germany. This means that we delete about 15,000 posts per month because of hate speech in Germany.” Richard Allan, We Are Working Hard To Fight Hate Speech and Have Already Made Great Progress, Facebook (June 19, 2017),

[10] Id. at 1302 (quoting OLG [Court of Appeals for Selected Matters], NJW, 38 (1985), 1720 (F.R.G.)).

[11] James Q. Whitman, Enforcing Civility and Respect: Three Societies, Yale L. Sch. Fac. Scholarship Series 1293, Although frequently used, the German Criminal Code provides little guidance on the meaning and scope of “insult” under the law. Section 185 of the Code simply states that “[i]nsult is punished by imprisonment for a term of up to one year or by a fine, and where the insult is made by means of physical assault, by a term of up to two years or by a fine.” See Strafgesetzbuch[StGB] [Penal Code], supra note 2, § 185. A German court of appeals distinguished insult from defamation by stressing that criminal insult does not need to have been communicated to a third party. Whitman, supra, at 1303 (internal citation and quotation marks omitted). That said, German courts have attempted to limit the law of insult to intentionally demeaning comments. Rudeness or rejection is not insulting per se. Id. at 1304. Thus, “expressions of hostility,” such as barring a non-German from a bar, “qualify as an insult only if they are understood as meaning that the owner of the establishment regards the targeted persons as unworthy of being served by him.” Id.

[12] Id. at 1296.

[13] Id. at 1297.

[14] Id. at 1306 n.70.

[15] David Elder, Defamation: A Lawyer’s Guide, Ch. 1, § 1:7 (West 2016).

[16] Ground Zero Museum Workshop v. Wilson, 813 F. Supp. 2d 678, 701 (D. Md. 2011).

[17] Franklin v. Pepco Holdings, Inc., 875 F. Supp. 2d 66, 74-75 (D.D.C. 2012).

[18] Elder, supra note 17, Ch. 1, § 1:7; see also Mitchell v. Random House, Inc., 865 F.2d 664, 670 (5th Cir. 1989) (finding that a book depicting the plaintiff in an unflattering manner was not defamatory, even if it could “reasonably be inferred [from the book] that [the plaintiff] was somewhat insensitive, unsympathetic, or callous” because “such [inferences] do[] not suffice for defamation under Mississippi law”).

[19] Veilleux v. Nat’l Broadcast. Co., 206 F.3d 92, 108 (1st Cir. 2000) (emphasis added).

[20] Scott Griffen, Defamation and Insult Laws in the OSCE Region: A Comparative Study, OSCE Representative on Freedom of the Media 103 (2017),

[21] Id.

[22] Whitman, supra note 10, at 1298.

[23] Id. at 300.

[24] See id. at 1306 n.74.

[25] Id.