The ongoing debate over regulating digital content reached a critical inflection point in India. This challenge is grave, as the proliferation of misinformation, amplified by electronic media, digital platforms, and social media, presents a serious threat to public order and democratic processes globally. The conventional safeguards of editorial control have become almost non-existent, largely because almost every individual has become a purveyor of news/information.
To counter misinformation specifically pertaining to the Government of India (“GoI”), the Press Information Bureau (“PIB”), operating under the Ministry of Information & Broadcasting (“MIB”), had established a Fact Check Unit (“FCU”) in November 2019. This mechanism verify information against authentic government open-source information, focuses on creating public awareness by identifying and flagging false information and posting corrections.
As of March 2025, the FCU identified more than 1500 instances of fake news in last three years. However, to seek a more streamlined control over online narratives, the Central Government amended the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (IT Rules 2021), in 2023 (“IT Rules”), and established a statutory FCU. It is tasked with identifying content relating to the “business of the Central Government”, and classify it as fake, false, or misleading. The critical component of this legal architecture was that if an intermediary, under Rule 3(1)(b) (“the Rule”), failed to remove this FCU-identified content, and thus failed to comply with the statutory due-diligence, it resulted in the immediate loss of statutory safe harbour protection under Section 79 of the IT Act, 2000, making intermediary liable for punishment under any law, as provided under Rule 7 of the IT Rules.
This unprecedented expansion of executive power was immediately challenged in the Bombay High Court’s (“BHC”) Kunal Kamra (“the case”) by comedians and press bodies like the Editors Guild of India. The earlier split verdict was then referred to a third judge, and then the majority ultimately struck down the Rule. Out of the many reasons, one of the rationales for declaring the Rule as unconstitutional was generation of a direct chilling effect on free speech of the citizen, by compelling the intermediary to take down the content flagged by FCU as fake news, or else it would immediately lose safe harbour protection. Hence, in a way, the intermediaries are facing, as described by Justice Patel, “Hobson’s choice” which, in this case, involves a choice of either losing one’s safe harbour by letting user’s post everything or prioritise risk-avoidance to protect their own business by removing the content the moment it gets published. Thus, the intermediaries will choose “throwing the user content under the bus than having the bus run over the entire business.”
Thus, the author attempts to examine how both the majority and the dissenting judges interpret the Rule in Kunal Kamra. The article shows that each judgment starts from an assumed meaning of the Rule, instead of first clarifying what the Rule legally requires. The majority treats “not to host” as a strict takedown mandate, while the dissent treats “reasonable efforts” as flexible enough to include disclaimers. The author argues that both readings are flawed because neither judge addresses the internal ambiguity and poor drafting of the Rule itself.
The Judicial Debate on the ‘Disclaimer’
One of the tests of reasonability under Article 19(2) is that of prongs of proportionality, which includes restriction by least restrictive means (“LRM”) to achieve the legitimate aim, as laid down in Modern Dental College, and was further solidified and applied in Anuradha Bhasin. The doctrine of LRM is activated in this case because the Petitioners argued that because the PIB was already functional on social media by issuing clarifications and corrections without the power to compel content removal, the creation of a statutory FCU was unnecessary. Further, the State failed to justify why it needed to replace a functional, non-restrictive awareness model with a coercive mandate for content excision. The LRM requires authorities to assess the existence of any alternative mechanism in furtherance of the legitimate goal. The Rule, however, introduced a mechanism that arguably failed this proportionality test by imposing an overwhelming and coercive requirement for content removal. This mechanism led directly to the central judicial debate in the case regarding the viability of a “disclaimer” of the fact that the post has been flagged by the FCU as fake news, as a less intrusive alternative to mandatory takedown under the Rule.
The Dissenting View: Disclaimer as “Reasonable Effort”
Justice Neela Gokhale, in her dissenting opinion, upheld the Rule by suggesting that issuing a disclaimer would constitute a sufficient safeguard and satisfy the least restrictive measure’s prong of proportionality. She argued that the Rule only required intermediaries to make reasonable efforts not to host content. In her view, issuing a disclaimer that flagged the content as FCU-identified would constitute a sufficient reasonable effort, and thus allowing the intermediary to retain safe harbour without necessarily taking down the content. She concluded that the consequences were not as severe as claimed because the option of issuance of a disclaimer was inherent in the Rule, and thus the provision did not suffer from manifest arbitrariness. Justice Gokhale’s stance was that this interpretation prevented the Rule from pre-empting the option of issuing a disclaimer.
However, this line of reasoning entails several logical leaps Primarily, it assumes that the mere presence of a disclaimer mitigates the harm the rule seeks to prevent without assessing whether such a measure actually mitigates impact of misinformation. Further, it treats the statute’s obligation to cause users not to host content on the same pedestal as merely labelling hosted content, without engaging with the text’s peculiar language that obliges intermediaries to influence the state of hosting itself.
Why the ‘Disclaimer’ was rejected by the Majority: The Compulsion of “Not to Host”
The result of the majority opinion directly rejects the argument that the Rule could be saved by reading in the option of a disclaimer. While both Justices Patel and Chandurkar ultimately reached a consensus in striking down the Rule, their interpretations of the intermediary’s obligations contained distinct judicial nuances. On the interpretation of “reasonable efforts,” Justice Patel reasoned that the phrase does not offer a choice between takedown and a disclaimer, as failure to remove FCU-flagged content would mean the intermediary failed to make such efforts altogether, and thus cause “functional contradiction” with the statutory obligation as it would still keep hosting the flagged content with the label of disclaimer. Justice Chandurkar concurred that “reasonable efforts” could not be read down to include a disclaimer, specifically rejecting the Union’s argument that this alternative was inherent in the Rule.
Further, Justice Patel viewed “not to host” as a strict command for the absolute excision of content, arguing that continuing to host with a disclaimer was not contemplated and would require “reading out” the words “not to host” from the statute. Justice Chandurkar provided a different statutory barrier, reasoning that an intermediary could not use a disclaimer because Rule 3(1)(b) strictly forbids them from “modifying” information; he held that appending a disclaimer would constitute an unauthorised modification of user content.
Analysis
While both Justices Patel and Chandurkar formed the majority that invalidated the Rule, they arrived at this conclusion via different interpretive routes regarding the intermediary’s options.
Justice Patel invalidated the Rule because it interpreted the phrase “to not host” as a mandate for the absolute excision of content from the public domain. This interpretation led directly to the finding of a Hobson’s choice, where intermediaries, fearing the loss of safe harbour protection, would inevitably choose unilateral removal over any form of debate or dissent. This reasoning and absolute excision was agreed and upheld by Justice Chandurkar. However, this is not a very welcoming interpretation of a law in the dynamic area like technology. Such interpretation of “to not host” is a classic example of overreach. It is based on a static view of digital data, where the presumption is that the “cause users not to host” means a pro-active obligation on the intermediary to compulsorily take down the post after it is publicly posted. However, had it been the case, such an interpretation would have rendered the phrase “reasonable efforts” altogether redundant. It is a settled principle of law that statutes should be interpreted to give effect to all words wherever possible. Hence, it is quite evident that takedown is not the only method of compliance. Further, Justice Patel reads the phrase “not to host” in isolation. A holistic reading of the provision makes it evident that it structurally separates the roles of the intermediary and the user. The intermediary’s obligation is confined to making “reasonable efforts”, while the duty “not to host” is placed squarely upon the user. The two are connected through a cause-effect relationship, where the intermediary’s reasonable efforts are intended to influence user conduct, not to substitute the intermediary’s judgment for that of the user by exercising direct control over publication itself. The phrase “make reasonable efforts” imports a qualitative, context-sensitive standard of conduct, and not an absolute command to eradicate all existing speech as a matter of rule. In cases where the post is causing situations like riots, or anticipates causing such situation, in the rarest of rare cases, the ‘reasonable’ effort having inherently proportionate and context-sensitive standard of conduct, may, by the order of the Court or that of the Government, extend up to calling off such a post. This is because under Section 79 of the IT Act, 2000, intermediaries lose immunity only if, after receiving “actual knowledge” (i.e. a court order or Section 69A government order) of the content, they fail to remove it. This exclusion was upheld in Shreya Singhal. However, the Rule seeks to by-pass this safeguard by empowering a government-appointed FCU to compel removals. If “to not host” is read as a strict takedown command, it in effect authorizes censorship on mere executive fiat, contrary to the stricter scope of Shreya Singhal.
On the other hand, Justice Chandurkar erred in holding that appending a disclaimer constitutes an unauthorised “modification” of information, Justice Chandurkar conflated the user’s prohibited conduct with the intermediary’s tools for compliance. While Rule 3(1)(b) instructs intermediaries to make reasonable efforts to ensure users do not “modify” prohibited content, this should not be interpreted to bar the intermediary itself from using corrective labels or disclaimers as part of its qualitative “reasonable efforts” to mitigate harm.
Way Forward
Instead of viewing not to host as a mandate for absolute excision, the phrase “make reasonable efforts” should be interpreted as a context-sensitive standard of conduct that focuses on influencing user behaviour rather than substituting the intermediary’s judgment for the user’s through direct control over publication. This interpretation must place the duty “not to host” squarely on the user, and the intermediary’s liability limits up to taking reasonable efforts. Moreover, to align with the Shreya Singhal, content removal should only be compelled through a court order or a Section 69A government order. Furthermore, as the current Rule remains “internally conflicted” and poorly drafted, the ultimate solution lies in legislative clarification to define “reasonable efforts” and specify permissible modes of compliance that is exactly in the middle of a total takedown and a simple disclaimer.
Conclusion
Hence, the reasoning of Justice Patel and Justice Chandurkar is unduly rigid and conceptually flawed. It adopts a static understanding of hosting, and also side-lines possible less restrictive alternatives. The reasoning given by Justice Neela Gokhale, even though qualifies the test of proportionality, fails to squarely fit in the Rule as the Rule strictly requires such ‘reasonable efforts’ that cause ‘users to not to host’ such information. The reasonable efforts of disclaimer, even though is ‘reasonable enough’ does not meet the threshold of the Rule. Hence, merely giving a disclaimer does not duly comply with the Rule. Thus, the Rule requires such a ‘reasonable effort’ that sits exactly in the middle of the extreme step like taking down the post altogether, and effort like putting a disclaimer.
Thus, the core failure lies in the drafting of the Rule, which leaves “reasonable efforts” and “not to host” undefined and internally conflicted. Judicial interpretation alone cannot cure this defect. A legislative clarification is required to specify permissible modes of compliance.
(This post has been authored by Shrushti Mahesh Taori, 5th-year B.A., LL.B. student at Maharashtra National Law University, Nagpur)
CITE AS: Shrushti Mahesh Taori, ‘Intermediary Liability, Judicial Interpretation, And The Cost Of An Unclear Rule’ (The Contemporary Law Forum, 24 February 2026) <https://tclf.in/2026/02/24/intermediary-liability-judicial-interpretation-and-the-cost-of-an-unclear-rule/> date of access.