On 9 October 2026, the Ministry of Home Affairs ordered mobile internet suspended within 4 km of the Janpath–Kartavya Path crossing in New Delhi, from 10 pm that night to 10 pm on 10 October. MediaNama reported that the notice was marked "top secret" and addressed to the nodal officers of Jio, Airtel, Vodafone Idea and MTNL. It states no grounds, does not explain the radius, and does not name the protest it relates to. The ministry did not publish it; the news agency ANI did.
The strongest case for the government
The state has a real argument. A crowd gathering without police permission, in a district under prohibitory orders, can turn dangerous quickly. Officials may fear that rumours or calls to mobilise spread faster than police can respond. Section 20(2) of the Telecommunications Act, 2023 gives the government this power on a public emergency or in the interest of public safety, where it is necessary or expedient for public order or to prevent incitement to an offence. A short, geographically bounded suspension is a milder tool than the multi-day, district-wide blackouts India has used before. A government could reasonably say this is the targeted measure courts asked for.
But the order's design is the problem, not the existence of the power.
What the law actually requires
The statute itself says suspension orders must be made "for reasons to be recorded in writing" (Section 20(2), Telecommunications Act, 2023). The Telecommunications (Temporary Suspension of Services) Rules, notified on 22 November 2024, add detail. According to SCC Online's summary, orders must be limited to the specific reasons for the suspension and must define the geographic area and services affected. They cannot exceed 15 days, and they must be published with their justification. A review committee is to meet within five days of the order. The Internet Freedom Foundation's first read adds that orders may issue only when the Section 20(2) objectives cannot be met by other reasonable means.
Measured against that, the 9 October order is hard to defend. A notice with no stated grounds cannot show that the "last resort" test was applied. A "top secret" classification is hard to square with a publication requirement. A radius with no explanation cannot show that the area was limited to the identified risk.
Escalation without explanation
The earlier shutdown shows what a compliant order looks like, and how far the second one departs from it. On 2 October the ministry suspended mobile internet for 12 hours, 9 am to 9 pm, within 1.5 km of Jantar Mantar. According to MediaNama, that order cited public safety, public order, and the prevention of a public emergency and incitement to offences. It invoked Section 20(2)(b) read with the 2024 Rules and recorded the Union Home Secretary's approval. It was obtained by MediaNama only on 6 October, four days after the event. Around 700 protesters were detained that day, per a PTI report cited by MediaNama.
The new order doubles the duration to 24 hours and widens the radius from 1.5 km to 4 km, a factor of about 2.7. Because area scales with the square of the radius, the covered area is roughly seven times larger. It also takes effect the same night it is dated. Residents, businesses and journalists in central Delhi had no notice, and no published basis to challenge. The government has offered no public evidence that a larger and longer measure was needed.
The Anuradha Bhasin standard
In Anuradha Bhasin v. Union of India (2020), the Supreme Court held that internet restrictions must meet a proportionality test: a legitimate aim, a rational connection to it, consideration of less restrictive alternatives, and a balance of impacts. It also held that restriction orders cannot be kept secret, and that they must be limited in time and territory and supported by sufficient factual material. Mere apprehension is not enough. An order that offers no facts cannot be tested against any of those stages. The 2024 Rules codified part of that reasoning but, as IFF notes, do not require a proportionality review and do not require publishing review-committee decisions.
Why this matters beyond one protest
The review mechanism has a structural weakness that this episode exposes. A committee that meets within five days of an order that lasts 24 hours reviews a shutdown that has already ended. IFF has argued that this makes review largely ineffective for the short suspensions that are most common. Short, repeated orders are also not capped cumulatively; the 15-day limit applies to each order. Two orders in eight days, each escalating, show how that works.
There is a practical cost too. Mobile internet in central Delhi carries payments, ride-hailing, navigation, delivery work and emergency communications. A 4-km radius around Janpath and Kartavya Path covers offices, hotels and transport hubs, not just a protest site. Blanket suspension also cuts off the people documenting events, which is often the point of such a measure. For a government that promotes digital public infrastructure as a national strength, switching it off in the capital with no stated reason damages that case.
What proportionate practice looks like
None of this requires denying the state its tools. It requires using the process it already wrote. Publish the order, and its reasons, when it is issued or soon after. State the facts that justify the area and duration, and explain why narrower measures would not do. Place the order before the review committee within the stated window and release its findings. Report aggregate figures so the public can see whether shutdowns are becoming routine.
If the ministry has grounds for a 4-km blackout, stating them costs nothing. If it does not, the order should not stand. Either way, an unreasoned, unpublished order contradicts both the statute and the court's ruling, and the next one will be harder to challenge unless this one is publicly examined.