In Regulating Free Speech in a Digital Age, David Bromell successfully combines both practical approaches and the application of international human rights law to trace the complexity of ‘hate speech’ regulations and reveals their implications for well-balanced policymaking. As an advocate of an open society where value-conflicts and even offending expressions are recognised as constituting elements of pluralistic democracy, Bromell sheds critical light on the concept of ‘hate speech’, which he deems ‘imprecise and misleading’ as it encompasses extremist and hateful, yet lawful speech (151).

His arguments are presented over ten chapters, divided into Part I (Regulating Harmful Digital Communication) and Part II (Hate, Harm and the Limits of Censorship). Part I provides a detailed explanation of the nature of digital platforms and their impact on democratic opinion formation. The principle of attention economy and the lack of transparency of algorithms damage democratic discourse by amplifying information that aligns with our existing beliefs (Chapter 3). Through an examination of the content moderation policies of Facebook and Twitter (now renamed ‘X’), Bromell emphasises the limits of self-regulation focused on the removal of content and de-platforming, turning readers’ attention to the necessity of protecting ‘borderline, but legal extremist material’ (63). Drawing on several examples including the de-platforming of former President Trump by Twitter in January 2021, Bromell sharply criticises the lack of democratic legitimacy of private digital platforms and their oversight boards in outlining the scope of allowed speech (Chapter 4). He argues for the necessity of a multi-stakeholder approach where governmental regulation, common industry standards and monitoring by civil society each play a key role.

Part II addresses the concrete challenges of policymaking in reaching a balance between the need to mitigate the actual harm caused by problematic digital communication and the need to guarantee the fundamental right to freedom of expression according to standards of international human rights law. Chapter 5 is a comparative legal analysis of ‘hate speech’ regulations in selected jurisdictions and an evaluation of existing approaches. In noting how the French Constitutional Court struck down several provisions that resembled provisions of Germany's Network Enforcement Act, Bromell implies that legal obligations on digital platforms to remove flagged content are always accompanied by dangers of private platforms abusing their power. Against this background, Bromell calls for scrupulous distinctions to be drawn when considering penalising harmful speech (Chapter 6). One possible approach is the Rabat Plan of Action adopted by the Office of the High Commissioner for Human Rights, which proposes a six-part threshold test (composed of the following six criteria: context, speaker, intent, content and form, extent of the speech act, and likelihood of harm) for defining expressions as criminal offences. Drawing on the principle of proportionality as a key international human rights standard, Bromell firmly concludes that ‘only incitement to discrimination, hostility or violence that meets all six criteria in the Rabat Plan of Action should be criminalised’ (142).

Indeed, a crucial challenge for human rights law in ‘hate speech’ regulation is the dilemma caused by the over-inclusion of protected characteristics. Bromell reasons that if ‘hate speech’ bans were to include all characteristics that deserve some kind of protection, it would not only involve ‘draconian limitations on fundamental rights of speech’ but also lead to enforcement based on subjective standards of complainants and authorities (173).

Despite the persuasiveness of the argumentation, it must be pointed out that it does give rise to uncertainty. Bromell's central argument is that it is public communication which incites discrimination, active hostility or violence that must be prohibited, not the manifestation of hatred and animosity. The criminal sanction for speech offences, according to this view, may only be justified when the speech act imminently causes harm such as incitement of active hostility and discrimination. Although this approach is based on the clear distinction between harmful effects and hatred as emotion or motivation, the problems that arise in determining the degree to which a speech act can be deemed as causing incitement as well as the practical issue of weighing the likelihood of risk remain unclear. It would have been helpful, for example, to illustrate how the six-part threshold test could be applied in some actual cases. Similarly, there is an inconsistency regarding Bromell's distinction between public and private communication in defining harmful speech. In his view, communication is harmful when it is ‘intended or likely to become open to witness’ (183). At the same time, he states that digital platforms significantly blur the boundaries between private and public (36–7). It is thus difficult to imagine how this—at least seemingly paradoxical—distinction could be implemented in policymaking.

This book is a valuable contribution for policymakers and legal experts dealing with speech regulation. Bromell's clear message, that the role of governments in countering the harmful consequences of problematic speech extends beyond the mere implementation of coercive measures, should be taken seriously. If not all instances of ‘hate speech’ should be prohibited, there is indeed a compelling need to focus on the communicative capacity of governments in promoting counter-speech and de-radicalisation efforts.