Decoding Intermediary Liability and the Chilling Effect on Online Speech

Introduction

Social media intermediaries and digital platforms have become the modern public square. People use them to relay all sorts of information about all sorts of things. As the volume of online content sharing has grown exponentially, the Indian government has steadily increased the obligations imposed upon intermediaries, the primary mediums for digital expression. It can be discernibly traced back to the enactment of the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (“Intermediary Guidelines”). Since then, the legal framework has seen various additions to it, both at the central and state levels. As this framework starts to flesh itself out, important questions must be raised with respect to it, such as: how fair is the law to the necessary stakeholders; how efficient is the law in achieving its goal of regulating harmful information online; and lastly, does the law align itself with the grundnorm, i.e. the Constitution. This article will attempt to analyse the following parameters with respect to the relevant legal framework. The objective here is to understand whether the government’s intention to ensure a safe online space is doing more harm than good?

Relevant Legal Framework

It is important to first establish the scope of the ‘relevant legal framework’. Section 69A of the Information Technology Act 2000 (“the Act”) authorises officers of the Central Government to direct any intermediary to block information generated and transmitted on its platform, based on certain reasonable restrictions. Section 79 allows safe harbour for these platforms, conditional on the fulfilment of due diligence obligations and other guidelines of the Central Government. They must remove access to information hosted on their platform, on receipt of actual knowledge of such data facilitating the commission of unlawful acts. Rule 3(1) of the Intermediary Guidelines specifies the due diligence obligations. Users must not transmit any information that falls within the categories enumerated under Rule 3(1)(b). They must be periodically informed of their obligations under the user agreement and that non-compliance may lead to the removal of content or even termination of access to the platform. Intermediaries must undertake ‘reasonable efforts’ on their own to make sure content is not infringing. Under Rule 3(1)(d), the platform must comply with orders to remove information violating the reasonable restrictions imposed under Article 19(2) of the Constitution. Takedown of such information must be done within three hours of being notified. The prescribed timeline for non-consensual intimate imagery (“NCII”) is two hours.  

A grievance redressal mechanism must be established to ensure a structured system for users to address their complaints. The name of the Grievance Officer and the procedure to lodge complaints must be prominently displayed. Rule 3A also establishes a Grievance Appellate Committee for any objection against decisions of the Grievance Officer. The guidelines define a Significant Social Media Intermediary as a platform enabling online interaction amongst more than 50 lakh registered users. These must observe additional due diligence obligations under Rule 4. For instance, they must deploy technical measures to proactively identify information related to rape, child sexual abuse, whether explicit or implicit. Such tools must undergo periodic reviews with regard to their fairness, security, and propensity for bias.

Part III of the Intermediary Guidelines applies to publishers of news and current affairs content as well as publishers of online curated content (like Netflix, Hotstar, etc.). However, the proviso to Rule 8(1) clarifies that Rules 15 and 16 under this Part shall also apply to intermediaries. Rule 15 gives the Ministry of Information and Broadcasting the power to direct intermediaries to block certain content on the basis of recommendations made by the Inter-Departmental Committee (“IDC”). Rule 16 gives the Ministry powers to issue interim orders in emergent situations. However, this framework under the guidelines has been stayed by the Bombay High Court, whose decision was subsequently reaffirmed by an interim order of the Madras High Court as having a nationwide effect. These cases are now consolidated before the Delhi High Court (with interim orders remaining in effect), pending final decision.  

The 2026 Amendment to the Intermediary Guidelines introduces obligations in relation to synthetically generated information (“SGI”), which is audio/visual information that is artificially created but appears to be authentic. The definition excludes information created through routine editing or good-faith creation that does not materially alter or falsify the data. The existing due diligence requirements will now extend to SGI with additional requirements such as proper labelling of such content. Platforms must deploy appropriate technical measures to prevent their generative tools from creating SGI restricted under Rule 3(3)(a)(i).

Fairness of the Framework

Powers that are excessive, vague or allow uncontrolled discretion are arbitrary and violate Article 14. The actual knowledge standard under S79(3)(b) of the Act and as elucidated in the Shreya Singhal judgment, is the current benchmark for content removal. This standard is met when the platform receives a court order or a notification from the appropriate government. The court stressed this standard to prevent intermediaries from becoming independent adjudicators of free speech. However, it has since been diluted. The 2023 Amendment introduced a new, vague standard of ‘reasonable efforts’, requiring intermediaries to exercise independent judgment. A similar obligation appears under Rule 3(1)(cb), where platforms must act not only upon receiving actual knowledge but also on ‘its own accord’. In the absence of a clear definition of ‘reasonable efforts’, intermediaries cannot decipher the extent of their obligations. An unreasonable burden is imposed, increasing the risk of curtailment of freedom of speech. A clear example emerged during the India-Pakistan conflict in May 2025, following the Pahalgam terror attack. The government reportedly ordered the blocking of over  8,000 accounts, including those of international news organisations. X itself stated that although it disagreed with many of the demands, it complied because of legal pressure. This further shows that when intermediaries fear the loss of safe harbour protection, they may over-censor content to avoid consequences. Vague, discretionary powers create a chilling effect on the right to speak freely and suppress criticism. These concerns surfaced in the Kunal Kamra case, in which the petitioner raised a complaint against the fact-check units. Although the issue surrounding ‘reasonable efforts’ was also raised, it did not form a part of the majority opinion. Justice Gokhale’s dissent is now the primary authority on this issue, which states that such efforts may arise when platforms become aware of unlawful content suo-moto, indicating a proactive obligation. However, it still failed to clearly define the standard or reconcile it with the actual knowledge benchmark. An appeal against the decision remains pending before the Supreme Court.

In X Corp. v Union of India, the Sahyog portal – a centralised system through which ministries, state departments, and police authorities can issue takedown orders under S79(3)(b) of the Act, was challenged. X argued that the government was creating a parallel blocking regime outside S69A. This concern deserves serious consideration. S79 orders may not carry the same binding force as S69A orders, yet they remain equally effective because non-compliance risks the loss of safe harbour. More importantly, these requests are opaque and lack safeguards like hearings and independent reviews. Thousands of police officials and state agencies can trigger a takedown under Rule 3(1)(d). As these powers move further down the administrative chain, oversight weakens and the possibility of misuse increases. The recent amendments proposed to the guidelines introduce Rule 3(4), which requires platforms to comply with any clarification, advisory, direction, or standard operating procedure of the Ministry. This allows the government to shape obligations in real-time, without formal rulemaking, notice and consultation. Particularly, if S79 is interpreted as granting the government independent powers, S69A becomes largely redundant. So, these must be interpreted harmoniously and in line with the maxim ut res magis valeat quam pereat, which urges an interpretation that does not render any provision void. Clearly, the government is circumventing important safeguards under S69A to expand its discretionary powers. However, the Karnataka High Court upheld the legal validity of the portal. An appeal before the Supreme Court remains pending. These instances demonstrate the arbitrary and unfair nature of the system towards both intermediaries and users.

Effectiveness of the Framework

Obligations imposed remain justified only when they can realistically achieve their intended objective. The current compliance timeline of three hours is so short that it leaves little room for platforms to review orders or even provide users an opportunity to be heard. It promotes a ‘comply first, question later’ approach that resembles aggressive censorship more than effective regulation. For instance, X follows a structured multi-step process for takedown orders, but such a short window leaves little scope for this due process to operate. Certain categories, such as NCII, naturally require urgent removal because delayed action may cause irreversible harm. Reasonably shorter timelines may be justified for such sensitive categories, but not when they apply broadly to all takedown requests. Obligations contemplated under these guidelines also require significant investment in infrastructure and manpower, to the extent that compliance might be costlier than non-compliance. Hence, these obligations may fail to achieve the law’s intended goal.

Moreover, safe harbour is not a privilege granted or withdrawn by the government at will, but a legal defence available when the intermediary faces liability in relation to unlawful information. Consequently, making platform governance conditional on safe harbour is ineffective, since non-compliance on their part becomes relevant only when legal liability actually arises. So, users possess no clear, independent remedy in case of a failure to follow important safeguards prescribed under the due diligence obligations. Without an independent enforcement mechanism, platforms have little incentive to comply consistently and may pick and choose the obligations they need to fulfil. As a result, the current framework cannot effectively achieve its objective of regulating the digital space.

Constitutional Concerns under the Framework

As stated before, arbitrary actions under these guidelines are a direct violation of Article 14. Unchecked powers lead to unreasonable restrictions on free speech, directly violating Article 19. Finally, Article 13 prohibits the state from enacting any law that abridges these fundamental rights.

Moreover, S69A specifies that only the Central Government and its authorised officers can issue blocking orders. State governments and other government departments may send removal requests under Rules 6 and 7 of the Information Technology (Procedure and Safeguards for Blocking for Access of Information by Public) Rules, 2009 (“Blocking Rules”), which are examined by a Designated Officer before taking necessary action. No independent powers are contemplated under these provisions for issuing blocking orders. So, Intermediary Guidelines cannot be used to grant independent powers to authorities like police officials, in defiance of this scheme. Delegated legislation must remain within the scope and framework of the parent act, otherwise it is ultra vires and cannot be given any effect. When rules expand authority beyond Parliament’s delegation, the executive exercises legislative power without constitutional sanction, undermining core constitutional principles such as separation of powers.

Conclusion

It is undeniable that the digital space cannot remain unregulated. Additional obligations on intermediaries are necessary, but the government’s current approach raises serious concerns due to its arbitrary nature. A balance must be sought between content regulation and protection of users’ rights. The ‘reasonable efforts’ standard especially needs reconsideration. Any Expectation of proactive obligation must be preceded by clear definitions, since vague standards create uncertainty and ultimately have a chilling effect on free speech. For instance, the government can clarify that pre-existing obligations, like deploying technical tools to identify and remove NCII and SGI, would satisfy this requirement. These measures also do not require platforms themselves to determine the legality of broad categories of content. They can be strengthened through audits and transparency requirements that ensure accountability in their functioning. The government must also reconsider the growing use of S79 for blocking orders and platforms like the Sahyog Portal. Without an amendment to the law, the government cannot expand its scope for administrative convenience. Until then, the government should introduce stronger safeguards like an independent review mechanism, reasoned orders, and adequate time for intermediaries to assess the requests properly. Rather than granting powers across multiple authorities, a similar system to the Blocking Rules should be followed, where requests pass through a coordinated authority. Faster procedures may be prescribed for sensitive categories, but they must remain proportionate and subject to ex-post reviews. Finally, in the long run, the government should contemplate separating platform governance from safe harbour protections. There should be independent enforceable obligations, limiting S79 to content-related obligations (similar to the European Union’s Digital Services Act). In conclusion, digital platforms must remain spaces where people can freely express opinions without facing unreasonable restrictions.

Author- Trishna Agrawalla is a fifth year student at National Law University, Odisha.

Edit0r- Anirudh Sachin Bajaj & Tanu Mehta