Website blocking orders under Section 69A grew roughly a hundredfold in the decade to 2023, according to figures MeitY disclosed under the Right to Information Act. By 2025, that figure had reached 24,000. The shift is qualitative as well as numerical: from targeted, supervised intervention to routine, opaque executive control.

Section 69A carries a tension the Supreme Court never resolved. It permits blocking wherever the Central Government “is satisfied that it is necessary or expedient so to do” in the interest of grounds tracking Article 19(2), which permits reasonable restrictions on free speech on a closed list of grounds including the sovereignty and integrity of India, public order, and incitement to an offence. “Necessary” implies proportionality; “expedient” means administratively convenient, a markedly lower bar that Article 19(2) itself does not authorise. A merely expedient restriction cannot be a reasonable one, so “expedient” is either void to that extent or must be read as “necessary”. Shreya Singhal appears to have proceeded on the second reading without adopting it. Its reasoning speaks of necessity throughout, but the judgment nowhere interprets “expedient” or holds that the word must be read down, and a narrowing that never enters the operative holding binds no one. What the Court actually decided was that Section 69A survives because of its procedural safeguards. The provision the nodal officer applies is therefore unchanged, and it still permits blocking on grounds of convenience. That matters more now that those safeguards are being weakened.

Three changes in the March 2026 draft illustrate how. First, proposed Rule 3(4) would make a platform’s immunity under Section 79, which protects platforms from liability for content posted by users, conditional on compliance with MeitY advisories and directions that are not formal Section 69A orders. This contradicts Shreya Singhal, which held that platforms neither must act on informal government pressure nor lose immunity for refusing. Platforms facing the loss of legal protection will remove content rather than resist.

Second, the expansion of the IDC’s mandate is textual and easily missed. Rule 14(2) currently confines the Committee to “complaints or grievances” about Code of Ethics violations; the draft substitutes “matters”, including those referred by the Ministry of Information and Broadcasting, and Rule 14(5) does the same for the Committee’s examination and recommendations. As the Internet Freedom Foundation has observed, this incentive structure produces systematic over-removal, whatever the motive behind any individual decision. This turns the IDC from an independent check into a body activated at ministerial discretion.

Third, the retention provisions in Rules 3(1)(g) and 3(1)(h) would preserve indefinite archives of citizens’ online expression across overlapping statutes. Speech that is known to be stored and retrievable is speech that gets self-censored.

Blocking orders have reached Whether any of these blocks in fact rested on an Article 19(2) ground is unknowable, because Rule 16 of the Information Technology (procedure And Safeguards For Blocking For Access Of Information By Public) Rules, 2009 applies to every blocking request and the action taken on it.

Amnesty International has warned that the amendments also risk violating India’s obligations under international human rights law. Six journalist bodies demanded the draft’s withdrawal at the Press Club of India in April 2026.

The constitutional position established in Shreya Singhal was clear: executive content control is permissible only within procedural limits that make judicial oversight practically possible, not merely theoretically available. A blocking regime whose orders are confidential, whose grounds go undisclosed, and whose liability rules push platforms toward removing speech without legal warrant has already moved outside those limits. The 2026 amendments do not address that problem. They make it permanent.

The consequences for queer children are devastating. Children seldom have meaningful agency to resist the demands of their caregivers, churches, or communities, even when those demands inflict profound psychological harm. Now, not even a well-intentioned state like Colorado can shield vulnerable queer youth, many of whom will be compelled to absorb the normative messaging of conversion therapy. The ideology central to conversion therapy teaches children that they are broken and must be fixed—an injury that will scar them well into adulthood. Yet, the Court reduces the practice to little more than a benign conversation: “all Ms. Chiles does is speak with clients.” This conclusion is possible only by ignoring the real harms children experience during and after a therapist’s “speech.”

Chiles is the latest in a stream of decisions demonstrating the Court’s special sympathy for perceived victims of LGBTQ+ inclusion. In 2021’s Fulton v. City of Philadelphia, the Court ruled for a Catholic foster care agency that precluded the placement of children in homes with same-sex parents without hearing from any LGBTQ+ parent applicants who may have been turned away. The Court reasoned that Philadelphia’s anti-discrimination policies created a risk that the city might treat religious adherents less favorably than secular parties at some future time. The opinion never acknowledges how this outcome facilitates—indeed, requires—the humiliating exclusion of queer people from serving as foster parents, or harms children in need of loving foster homes that queer families can provide.

More recently, in 2025’s Mahmoud v. Taylor, the Court sided with parents challenging a school district’s policy denying parents opt-outs from LGBTQ-inclusive curricula. The majority opinion minimized the threat that opt-out rights pose to public education more generally, and it was silent about the dignitary, emotional, and stigmatic harms that queer children will experience when their peers exit a classroom before any educational acknowledgment of queer themes, characters, or stories. Instead, the opinion fixates on the religious burdens conservative parents will suffer when their children are exposed to the reality that LGBTQ+ people exist. According to the Court, these LGBTQ+ inclusive lessons impart intolerable normative lessons that gravely threaten parents’ religious beliefs. First, it is difficult to discern any limiting principle in Mahmoud; does public education not exist, at least in part, to transmit norms of civility and respect for others in a pluralistic society? Second, the Court fails to even consider the normative messages queer children will internalise when their identities are erased from the classroom.

In these and other recent decisions, like 2018’s Masterpiece Cakeshop granting a Christian baker the First Amendment right to deny services for a same-sex wedding, the Court appears myopically sensitive to the harms that may befall conservative religious adherents within a legal landscape that suddenly treats queer people as human beings worthy of inclusion and respect. In an ominous sign, the Court recently agreed to hear a case next term, St. Mary Catholic Parish v. Roy, offering yet another opportunity to weigh religious liberty claims against LGBTQ-inclusive policies. The question in St. Mary’s Catholic Parish centers on whether Colorado may constitutionally exclude a Catholic preschool from its universal preschool program because the preschool’s admission policy excludes LGBTQ children and children with LGBTQ parents. Given the Court’s recent pattern, it is easy to predict the likely outcome: a law designed to prevent LGBTQ exclusion will be twisted into a law that excludes religious adherents.

In the last decade, the Court has remade the Constitution in line with right-wing policy preferences. Today, harms to religious adherents’ commitment to LGBTQ exclusion garner maximum constitutional protection, while the resulting exclusionary harms to LGBTQ people receive none. Simply put, the contemporary Court’s interpretation of the First Amendment privileges religious intolerance over equality for all.

Applicable IHRL

Criminalisation of hate speech is governed by Articles 19 and 20 of the International Covenant on Civil and Political Rights (ICCPR), which India has ratified. The Rabat Plan of Action, developed under the auspices of the Office of the United Nations High Commissioner for Human Rights, provides guidance on the interpretation of these provisions.

Under Article 20(2) ICCPR, criminalisation is confined to intentional public advocacy of national, racial, or religious hatred that constitutes incitement to discrimination, hostility, or violence. The Rabat Plan emphasises that criminal law must be reserved for the most severe forms of expression, and that legislatures bear responsibility for narrowly tailoring offences to clearly separate protected speech from punishable advocacy. In guiding legislative design, the Rabat Plan draws on the Camden Principles, which require narrow definitions of “hatred” and “hostility” and limit criminal liability to inciting advocacy posing an imminent risk of harm. Finally, the Rabat Plan further recommends that enforcement be guided by a six-factor assessment.

Even where Article 20(2) is engaged, restrictions on speech must independently satisfy Article 19(3) ICCPR, including the requirements of legality, legitimacy, and necessity, as clarified by the Human Rights Committee.

The Bill’s misalignment with IHRL

Although framed as a criminal response to hate speech, the Bill does not conform to the constraints IHRL imposes on criminalising expression:

Firstly, the Bill conflates ‘hate speech’ and ‘hate crimes’ by defining offences primarily by reference to communication of expression. However, IHRL distinguishes these terms. Hate speech concerns inciting advocacy of national, racial, or religious hatred under Article 20(2) of the ICCPR and remains subject to Article 19 safeguards. Hate crimes, by contrast, involve bias-motivated criminal acts and are regulated through equality and non-discrimination guarantees.

Secondly, the Bill extends criminal liability beyond the closed categories recognised under Article 20(2) – namely national, racial, or religious hatred – by encompassing additional characteristics like sexual orientation, gender, place of birth, residence, and language. While these characteristics may warrant protection through other legal or policy measures, their inclusion within a criminal prohibition on speech exceeds the limited scope contemplated under Article 20(2).

Thirdly, the Bill lowers the threshold of hatred under Article 20(2) by criminalising expression intended to promote “ill-will”, “disharmony”, or “enmity”, with hatred operating as only one among several alternative triggers. The Bill also omits any definition of “hatred” or “hostility,” despite the Camden Principles requiring these terms to be narrowly understood as “intense and irrational emotions of opprobrium.”

Finally, the Bill dilutes the incitement standard by treating it as interchangeable with promotion, propagation, abetment, and attempts to generate ill-will or disharmony. This departs from the Rabat Plan and Camden Principles, which confine criminalisation to inciting advocacy creating a clear or reasonably foreseeable risk of discrimination, hostility, or violence, and caution against collapsing incitement into mere promotion.

These issues demonstrate the Bill’s failure to meet Article 20(2) ICCPR threshold requirements, leaving limited scope for a meaningful proportionality analysis under Article 19(3), ICCPR. Even if Article 19(3) were engaged, the Bill would, at a minimum, fail the legality requirement since it confers unconstrained discretionary preventive powers on LEAs. It also, arguably, fails the necessity prong, which requires speech-restrictive measures to constitute the least intrusive means of achieving a legitimate aim (like maintaining public order). This is because existing federal frameworks already enable State-level LEAs to address overlapping categories of electronic expression through less restrictive means, including directions issued to intermediaries under Rule 3(1)(d), 2021 IT Rules. This framework, though not without its infirmities, largely operates within constitutionally recognised limits on speech, and incorporates at least some procedural safeguards like reasoned intimations and periodic reviews. Further, comparable prohibitions already exist at the federal level, rendering the Bill’s duplicative criminalisation difficult to justify as necessary.

Conclusion

The Bill’s implications extend beyond Karnataka, particularly due to its application to electronic expression, which cannot be meaningfully ringfenced within State boundaries. Assessing the Bill against IHRL standards is therefore germane, especially considering the Supreme Court’s recognition in Vishaka and PUCL that IHRL obligations may inform legislative interpretation where fundamental rights are implicated. Measured against IHRL standards, this analysis finds the Bill deficient, warranting reconsideration before it is operationalised.

The Avia law required online platform operators that connect several people and facilitate sharing of content to take down any “manifestly” illegal content within 24 hours of being notified by anyone. Further, content concerning child pornography and terrorism was required to be taken down within 1 hour of notification. Non-compliance with these obligations would have resulted in fines amounting to hundreds of thousands of Euros and up to 4% of the global annual revenue of the platform in extreme situations.

The Constitutional Council emphasised the importance of freedom of expression and communication in a democratic set up. Accordingly, while it found the aim behind the sought restrictions to be legitimate, it held the restrictions to be disproportionate and not necessary. The primary reason behind this conclusion was the short duration prescribed for taking down of content, without giving an opportunity to seek prior opinion of a judge. This was considered important as they were required to decide on issues involving legal technicalities. This became even more difficult due to the possibility of numerous complaints being received by the platforms, and hefty penalties prescribed in case of non-compliance.

This decision, thus, deals with the fundamental balance to be struck between freedom of expression and the accompanying restrictions. Particularly in the French context, as noted by the EDRi, the French law concerning terrorism, including provocation and apology of terrorism, has been widely criticised for being vague and overbroad. It does not provide clarity as to the content that has to be censored. In such a scenario, assessment of content on these parameters within one hour would have set an unreasonable burden on the online platform. This is more so, as also noted by the court, when assessment of any content in light of a complex legal matrix within even a 24 hour deadline is considered a disproportionately high obligation.

Furthermore, the above onerous obligations, in effect, would have delegated censorship measures to be enforced by intermediaries, as warned against by UN Special Rapporteur Frank La Rue. Given the short timelines and high penalties, this would have led to a chilling effect on the platforms as they would tend to err on the side of caution. Accordingly, most platforms would have resorted to prior restraint and adopted automatic takedown mechanisms where content would be removed without conducting an adequate check.

A facet that the Constitutional Council did not touch upon is the impact of such prescriptions for small players or not-for-profit organisations. The burden is particularly high on them as they do not usually have the means to incorporate a 24/7 monitoring mechanism, either automatic or manual, which was in effect mandatory as per the Avia law. This was aggravated due to the presumption under the law of intent on the part of the platform in case proportional or necessary examination of the content is not carried out. Accordingly, they might have been carved out of the market due to the possible negative ramifications.

A reduction in public content sharing platforms would have further impinged upon the freedom of the public to exchange ideas and information. Some of the legitimate content that would have possibly then been curbed includes political dissent seeking accountability of the government. Therefore, it would have diminished the utility of intermediary platforms by lifting the cover of safe harbour and threatening them with penalties.

Hence, as nations across the world dabble with regulation of content on online platforms, many a time tending towards overregulation, this decision appears to be a silver lining. Hopefully, it would create a ripple impacting similar legislative attempts of regulatory overreach in other jurisdictions such as Germany that have been met with severe criticism for such proposals.