Legal philosopher Jeremy Waldron called hate speech a “slow-acting poison”. The phrase captures the way public hatred gathers force through repetition, political reward, institutional silence, and social normalisation. Its deepest injury lies in the weakening of equal civic standing: the assurance that every citizen can inhabit public life with dignity.

That is the lens through which the Supreme Court’s judgment in Ashwini Kumar Upadhyay v Union of India should be read. Justices Vikram Nath and Sandeep Mehta, who delivered it, were right to respect the limits of judicial power in criminal law. Penal offences require legislative text. The flaw lies in the court’s further conclusion that India’s existing statutory framework adequately deals with hate speech.

The judgment, delivered on April 29, held that it could neither create offences nor prescribe punishments through judicial directions. It also held that existing substantive and procedural law supplied adequate remedies, with failures arising from application and enforcement. In Brinda Karat’s appeal, the court clarified that a previous sanction is required only at the stage of taking cognisance, while registration of an FIR and investigation can proceed where information discloses a cognisable offence. Sanction had become a shield against investigation itself. By removing that threshold barrier, the court has made it harder for magistrates and police authorities to postpone action in serious cases.

But the court left the petitioners dissatisfied for other reasons. The amicus, Sanjay Hegde, submitted that the current legal framework against hate speech appears comprehensive on paper, while remaining structurally deficient because existing provisions are reactive and designed for isolated acts, instead of systemic or corporate dissemination. Other petitioners stressed the special danger of hate speech by persons holding authority, since official or political status gives such speech a semblance of legitimacy. The judgment records these submissions yet answers them mainly through the route of criminal procedure: approach senior police officers, approach a magistrate, and then approach constitutional courts.

That route is real. Its existence gives little comfort when the complaint concerns a powerful speaker, delayed registration, diluted charges, or a police decision to treat a civic equality problem as a minor public order matter. The petitioners were asking whether the law has the right categories. The court answered by listing the available doors.

Hate speech in India

India’s hate speech regime is made up of provisions created for different anxieties. Section 153A of the Penal Code dealt with the promotion of enmity between groups. Section 153B addressed assertions prejudicial to national integration. Section 295A punished deliberate and malicious religious insult. Section 505 dealt with rumours, alarms, and public mischief. Election law covers communal appeals. Media and technology laws touch films, broadcasting, cable television and online circulation in limited ways. These provisions now continue, with changes, under the Bharatiya Nyaya Sanhita (BNS) and allied regimes.

This legal abundance produces uneven results. A religious insult provision can travel quickly against satire, reformist critique or historical argument. A public tranquillity provision may miss the civic harm caused by repeated dehumanisation. Expressions such as ill-will and disharmony give wide discretion to the police. The same framework can be energetic against weak speakers and hesitant against influential ones.

The Centre for Communication Governance at National Law University Delhi mapped this field in detail. Its 2018 report said the proliferation of hate speech in India arises amid many laws regulating speech, and that judicial standards under these laws are often applied inconsistently. That finding should have made the court cautious before describing the framework as adequate. The presence of many laws may itself be part of the problem when they pull enforcement towards sentiment, order, morality, elections, security, and media control, with equality appearing only in fragments.

The Law Commission’s 267th Report took an important step by acknowledging that hate speech lacks a legal definition in India. It proposed new provisions on incitement to hatred and speech causing fear, alarm or provocation of violence on grounds including religion, race, caste, community, sex, gender identity, sexual orientation, place of birth, residence, language, disability, and tribe. The court noticed the report, yet gave insufficient weight to its diagnosis.

Academic Siddharth Narrain’s critique of that report sharpens the issue. Writing for the Socio-Legal Review Forum, he called it a “mixed bag”: valuable in its mapping of global jurisprudence, Indian law and the liberty-equality tension, yet narrow because it rested mainly on criminal law amendments. He warned that the proposed provisions could lower existing free speech protections and be used to curb legitimate speech and dissent, especially given the misuse of Sections 153A and 295A against creative expression and criticism.

Narrain’s insight prevents a simplistic call for more penal law. The choice before India is between two flawed instincts: satisfaction with scattered provisions and expansion of criminal offences without safeguards. The harder task is statutory design that targets civic harm while limiting official discretion. His more recent work frames the issue as a need to move away from over-reliance on public order and towards an anti-discrimination framework attentive to India’s structural and historical inequalities.

Narrain’s concern stems from the view that public order captures only part of the injury. A speech may create no immediate riot and still lower the civic status of a community. A boycott call, a recurring conspiracy narrative, a vilifying television campaign, or a Minister’s coded statement can alter social relations before violence arrives.

Legal philosopher Jeremy Waldron locates the harm of hate speech in the loss of “public assurance”: the confidence that citizens can live in society as equals, without being marked as outsiders or legitimate targets of hostility. This is why a hate speech law cannot be built only around public order or hurt sentiment. It must also account for the damage caused to equal civic standing.

Political theorist Bhikhu Parekh’s work similarly explains how hate speech lowers the social position of a group by presenting its members as legitimate objects of hostility. Its harm lies in the public construction of a target community as dangerous, inferior, or undeserving of equal regard. This makes discrimination appear reasonable and hostility appear justified.

Media’s role

Digital media intensifies the problem. Narrain’s work on social media and violence shows that, after the Supreme Court’s striking down of Section 66A of the Information Technology Act in the Shreya Singhal case in 2015, enforcement discourse shifted towards categories such as “objectionable” and “provocative” content. The focus moved from content and intention to pre-empting circulation. Monitoring labs, takedown mechanisms, keyword searches and platform architecture began to shape the category of hate speech. His 2024 Yale essay extends this through the idea of online propaganda, stressing virality, cross-platform circulation, encrypted messaging, algorithmic curation, and the role of state, sub-state, and private actors in amplification.

This is the context in which the Supreme Court’s procedural answer appears thin. A victim facing a coordinated digital campaign needs more than a route from the police station to the magistrate. A dissenter facing an overbroad complaint needs safeguards before the same machinery moves too quickly.

The Rabat Plan of Action offers a workable threshold. Developed within the UN human rights system for applying Article 20 of the International Covenant on Civil and Political Rights (ICCPR), it requires a high threshold for restrictions on expression. Its six factors are context, speaker status, intent, content and form, extent of dissemination, and likelihood of harm, including imminence. Context asks whether the speech occurs amid elections, communal tension or a history of violence. Speaker status asks whether the speaker is a Minister, party leader, cleric, broadcaster, platform account or private citizen. Extent measures reach, repetition and medium. Likelihood assesses the probability of discrimination, hostility or violence.

This test could have given the Court a principled middle path. It could have adopted the Rabat factors as interpretive guidance for police, magistrates, and High Courts while leaving legislation to Parliament. It could have required reasoned action or inaction in complaints involving public officials and influential political speakers. It could have sought enforcement data from States. It could have directed High Courts to frame time-bound practice directions for complaints where speaker authority, target vulnerability, and dissemination reach raise serious concerns. These measures would have structured enforcement while preserving the boundary against judicial law-making.

Comparative experience points in the same direction. Canada criminalises public incitement of hatred likely to lead to breach of peace and wilful promotion of hatred against identifiable groups, with statutory defences. South Africa’s Constitutional Court in Qwelane rejected the vague word “hurtful” and preserved a narrower approach tied to harm and incitement of hatred. The lesson for India is precision: protected groups must be identified, mental elements must be demanding, and safeguards for journalism, scholarship, artistic expression, religious reform, anti-caste critique, feminist critique, and political dissent must be explicit.

A sound Indian law should cover public advocacy of hatred against identifiable groups on protected grounds, including religion, caste, tribe, language, ethnicity, race, sex, gender identity, sexual orientation, disability, place of birth, residence, nationality, and migrant status. It should require intention to incite, or knowledge of a serious likelihood of inciting, discrimination, hostility, or violence. It should require decision-makers to assess speaker authority, audience, vulnerability, medium, repetition, reach, history of violence, and proximity to elections.

The statute should also grade consequences. Criminal prosecution should be reserved for threats, direct incitement, dehumanising calls, organised vilification and calls for social or economic exclusion, discrimination or violence. Other situations may call for takedown orders, rights of reply, broadcaster sanctions, election consequences, platform transparency duties, civil remedies and public counter-speech. Police refusal to act in serious complaints should carry brief written reasons and a time-bound magistrate review. Platform regulation should be transparent enough to avoid rule by opaque algorithms or informal State pressure.

The Supreme Court had no obligation to draft this statute. Its institutional task was smaller and still important: identify the inadequacy of the present framework, preserve the legislative domain, and give interpretive structure to enforcement. The final direction asking High Courts to consider practice directions is useful, yet modest. It leaves the central doctrinal gap untouched.

Restraint has value when it protects democratic law-making. It becomes costly when it converts structural failure into routine enforcement failure. India’s present law moves uncertainly among religious offence, public order, national integration, electoral morality, media regulation, and equality. That uncertainty gives the State discretion where precision is needed and hesitation where accountability is required.

Waldron’s metaphor reminds us that hate speech alters the civic atmosphere over time. Narrain’s work adds a second caution: the cure can become a method of control when vague categories, policing discretion and platform architecture replace doctrine. By calling the existing framework adequate, the Supreme Court has made legislative reform appear less urgent. That is the test the judgment failed.

V. Venkatesan is a senior legal journalist and a researcher

Also Read | Hate speech has moved from the margins to the mainstream of governance

Also Read | The new face of digital violence