Expanding Freedom and Democracy
Freedom on the Net 2017

India

Partly Free
59
100
A Obstacles to Access 13 25
B Limits on Content 26 35
C Violations of User Rights 20 40
Last Year's Score & Status
59 100 Partly Free
Scores are based on a scale of 0 (least free) to 100 (most free). See the research methodology and report acknowledgements.

header1 Key Developments, June 1, 2016 - May 31, 2017

  • Internet access and speeds improved (see Availability and Ease of Access: Key Indicators).
  • Local authorities ordered temporary telecommunication service shutdowns in at least 37 separate reported incidents (see Restrictions on Connectivity).
  • Officials ordered service providers in the Kashmir valley to block 22 social media sites for a month, including Facebook, Twitter and WhatsApp (see Blocking and Filtering).
  • Over 20 people were detained for online comments about religion or political issues ranging from a water dispute between Karnataka and Tamil Nadu to a demonetization policy intended to combat corruption; a Kashmiri was held for several weeks in Chhattisgarh for sharing an “anti-India” cartoon (see Prosecutions and Detentions for Online Activities).
  • The Supreme Court recognized privacy as a fundamental right in a landmark ruling in August 2017 (see Surveillance, Privacy and Anonymity).

header2 Introduction

Internet freedom remained stable in 2017 after a decline in 2016. Improving access was offset by network and social media shutdowns ordered by authorities.

The number of internet subscribers and internet penetration increased significantly during the reporting period, as India consolidated its position as the world’s second largest internet consumer base after China. Both governmental and nongovernmental entities made efforts to bridge the digital divide between urban and rural areas.

A Constitution Bench of the Supreme Court held that privacy is a fundamental right under the Indian Constitution and a committee was set up to frame a data protection framework for India.

However, other developments undermined internet freedom. The number of network shutdowns increased substantially and local authorities ordered service providers to temporarily shut down internet access in at least 37 reported incidents in various states.

There was also an increase in the number of criminal charges for online speech filed under the IT Act and provisions of the penal code. Many people were detained for content circulated on WhatsApp or published on Facebook, including group administrators who were not responsible for the content.

A Obstacles to Access

Internet penetration in India continued to increase in 2017 with mobile penetration playing a significant role. Inadequate infrastructure remains a significant obstacle to access, especially in rural areas; however, various governmental and nongovernmental efforts to improve access nationwide are underway. Nearly 40 information communication technology (ICT) shutdowns were ordered by local authorities, some lasting several months in Jammu and Kashmir. The top ten internet service providers (ISPs) still hold almost the entire market share, but strong competition among them continues.

Availability and Ease of Access

Internet access and speeds improved during the reporting period (see Key Access Indicators). India had the second largest number of Internet subscribers in the world after China in 2017, having overtaken the United States.1 Official statistics recorded over 431 million subscribers in June 2017,2 though only 21.6 million had fixed-line internet connections.3 There were an estimated 269 million internet users in urban India and 163 million in rural India in 2016.4

However, internet penetration remains low, reaching 33 percent in June 2017,5 up from 27 percent in June 2016.6 Mobile penetration was much higher, reaching 92 percent by June 2017,7 up from 81 percent the previous year.8 The Broadband Commission ranked India 78 out of 196 countries in terms of mobile broadband penetration,9 up from 156 out of 179 countries the previous year.10

While India’s average connection speed was one of the lowest in Asia,11 it is catching up to the global average, which Akamai documented at 7.2 Mbps in the first quarter of 2017.12 Approximately 34 percent of all internet users had narrowband subscriptions in 2016,13 down from 56 percent in 2015.14 Despite overall growth, India has a relatively low adoption rate for high speed broadband (faster than 10 Mbps), at just 19 percent,15 though this rate grew by 285 percent during the course of 2016.16 The minimum speed required to qualify as broadband in India has been 512 Kbps since 2012,17 though the Telecom Regulatory Authority of India (TRAI) has recommended raising the threshold to 2 Mbps.18

The Global Information Technology Report by the World Economic Forum and INSEAD ranked India in eighth place out of 139 countries for affordable internet access in 2016.19 It was previously in first place,20 and per minute cellular and fixed broadband tariffs are still among the lowest in the world.21 While the cheapest internet plans might seem extremely affordable with respect to the average monthly income, India has significant income inequality.22

India ranked 66 out of 137 countries for infrastructure in 2017, according to the World Economic Forum’s Global Competitiveness Index.23 Though up from 68 the previous year,24 the results suggest poor infrastructure is still an obstacle to access. India ranked a low 88 for electricity supply;25 and 110 for technological readiness, the capacity of a country to fully leverage ICTs in daily activities.26 Only 27 percent of all Indian schools had a computer in 2016.27 That increased to nearly 80 percent at secondary level and above,28 but less than half were connected to the internet.29

Public and private sector initiatives to improve access are underway. The government is developing free public Wi-Fi zones in major cities,30 with some operational in the past year.31 In January 2017, the Maharashtra government activated 500 Wi-Fi hotspots across the city of Mumbai,32 though further expansion fell short, and they were only free until August 2017.33 During the coverage period of this report, Google partnered with the public sector company RailTel to provide free Wi-Fi at train stations,34 connecting 100 by the end of 2016.35 Over 5 million people were using the service every month.36

The government’s Digital India Programme, launched in 201437 is expected to be implemented by 2018.38 It aims to connect India’s gram panchayats, institutions of self-government in rural areas, via fiber-optic cables,39 ensuring universal broadband access with accompanying e-literacy programs. Internet-connected common service centers (CSCs) aim to cover all 250,000 gram panchayats;40 as of March 2016, 157,000 had been established, with 20,000 operated by women.41 The program proposes to use satellites, balloons, or drones to push faster digital connections to remote parts of the country,42 as well as multiple system operators such as cable TV services, which already have last-mile connectivity.43 As a result of the program, electronic transactions related to e-governance projects almost doubled in 2015.44

Such initiatives took on new significance during the coverage period, which saw a major push to digitize financial transactions. The government demonetized currency notes in the denominations of INR 500 and INR 1000 (US$7.5 and $15) in November 2016; the notes made up over 85 percent of the total currency in circulation.45 A Digi Dhan Abhiyan program was designed to promote digital payments to more than ten million inhabitants of rural areas,46 reaching 2.5 million people by the end of the year.47 The Ministry of Electronics and Information Technology (MeitY) also announced an alliance with Google to raise awareness of digital security surrounding payments.48

Language remains a barrier to access. With 22 official languages, only about 12 percent of the population of India speaks English,49 yet more than half the content available online is in English,50 and over 100 languages were unrepresented online in 2013.51 Projects to encourage local language usage are underway. In 2014, the National Internet Exchange of India (NIXI), which operates and manages Indian domain names, launched the Dot Bharat domain for local language URLs.52 By April 2017, the number of local language users in India had overtaken the number who rely on English.53 One study showed that nearly 70 percent of Indian internet

users consider local language content to be more reliable than English content.54 In April 2017, Google partnered with a local business federation to develop content in Indic languages.55

Studies have shown that economic and social conditions result in barriers to internet access for women, and only 29 percent of Indian internet users were female in 2015.56 Internet usage was lower among rural women (25 percent), though it had grown by 30 percent since 2015.57 Twenty-four percent of Indian Facebook users were women, well below the global average of forty-four percent, according to one calculation.58 Internet Saathi, a partnership between Google and Tata Trusts to promote digital literacy among rural women, was active in 25,000 villages across 10 states by October 2016,59 training more than 500 participants a week.60

Restrictions on Connectivity

The Indian government does not routinely block the protocols or tools that allow for instant, person-to-person communication, although local authorities around India have restricted ICT connectivity and usage during times of perceived unrest since at least 2010.61

The frequency, geographic distribution, and duration of these shutdowns have increased significantly in the past three years. During the coverage period of this report, authorities ordered providers to restrict local mobile phone, SMS, wireless, and occasionally fixed-line internet service in at least 37 reported incidents, which lasted for hours, weeks, or even months at a stretch.62

Local authorities have justified these orders under Section 144 of the Code of Criminal Procedure (1973), which permits broad state action to curb any violation of law and order.63 The Gujarat High Court upheld the use of this general law to order shutdowns in September 2015.64 The Supreme Court is yet to consider the matter substantively and refused a petition challenging it in early 2016.65

Other laws used to justify shutdowns also lack specificity. Section 69A of the Information Technology (IT) Act, which permits the central government to order website blocks (see Limits on Content) has been considered

to apply to blocking of service. Section 5 of the Indian Telegraph Act, which allows state and central authorities to order that any message not be transmitted in public emergencies, has also been cited in support of service disruptions.66 State officials in Odisha suspended service for 48 hours under the Telegraph Act after content considered to derogate Hindu deities resulted in violence.67

In August 2017, outside the coverage period of this report, the Department of Telecommunications of the Central Government issued new rules under the Telegraph Act to regulate the temporary suspension of telecom services.68 The rules authorize national or state-level officials to issue temporary suspension orders to shut down telecommunications services in times of public emergency or threats to public safety.69

With at least 12 documented incidents, Jammu and Kashmir continued to be the most affected state. Shutdowns affected both mobile and fixed-line connections, and the longest lasted several months.70

  • In June 2016, mobile internet services were suspended across the state for three days after a temple was vandalized, launching an outbreak of violence.71 They were suspended for a day on a second occasion in the Jammu region because of security fears surrounding an annual wrestling contest hosted on contested land.72
  • In July 2016, security forces shot and killed militant commander Burhan Wani in Kashmir, sparking widespread protests.73 All mobile service providers except BSNL, the state operator, suspended phone service in the Kashmir valley, and all operators suspended mobile internet throughout the state. The phone services were restored after a few days. Mobile internet services were restored in the Jammu region after 17 days. In the Kashmir region, mobile internet for post-paid subscribers remained unavailable for 134 days. Internet was not restored for prepaid subscribers until January 2017, almost 6 months later.74 Broadband internet in the valley was also shut down for 5 days in August due to the ongoing tensions between protestors and security forces.75
  • In September 2016, broadband services across Kashmir were suspended for an additional five days prior to the Eid festival.76
  • In April 2017, both mobile and fixed-line broadband internet services were suspended for a few days in the Kashmir valley when local by-elections sparked unrest. The measure was intended to curb rumors, but had the opposite effect, reports said.77 Mobile internet across the valley was suspended again amid student protests.78 Social media applications were also blocked (See Blocking and Filtering).

Shutdowns were implemented in several more states, including Maharashtra,79 Bihar,80 Odisha,81 Uttar Pradesh,82 and Arunachal Pradesh.83 Haryana and Rajasthan saw at least seven incidents each.84 Haryana shutdowns came in response to ongoing, sometimes violent protests by the Jat caste over their eligibility for government affirmative action quotas.85 In Rajasthan, internet was blocked on at least four occasions in Bhilwara district, once following the murder of a Hindu nationalist activist in September 2016,86 and three times within two weeks in December 2016 after communal violence flared in December 2016.87

The government does not exert much control over the internet infrastructure. Twelve submarine cables connect India to the global internet;88 ten are consortium owned, while the others are private.89 There are gateways to the international internet in Chennai, Mumbai,90 and Agartala in Tripura, which facilitates connectivity in north-eastern states.91 There are four landing stations where the cables meet the mainland in Mumbai, and three in Chennai; Digha, Kochi and Tuticorin also have one cable landing station each.92 BSNL, the state-owned telecom operator, owns two of them; the rest are privately owned. Major telecom operators Bharti Airtel and Tata Communications own three stations each.93 These cable landing stations imposed hefty fees on ISPs until regulators mandated a reduction in 2013.94 Tata Communications and Airtel challenged that reduction in the Madras High Court. A single judge dismissed it, and an appeal was pending in early 2017.95

Undersea cables are mainstays of mobile and internet communications and any damage to them leads to service disruptions. In December 2016, Cyclone Vardah caused damage to Airtel’s undersea cable at Chennai, slowing internet speeds.96

Over 80 percent of telecommunications towers are privately owned.97 Market share is split between Indus Towers, a joint venture between Bharti Infratel, Vodafone, and Idea Cellular (31 percent); BSNL (18 percent); and Reliance Infratel (12 percent), and Bharti Infratel (10 percent) according to 2015 figures.98

ICT Market

There are 157 operational ISPs in India.99 While there is no monopoly, the top 10 ISPs control over 98 percent of the market.100 Thanks to inaugural promotional plans, Reliance Jio made massive gains to achieve the highest ISP market share of 29 percent in 2017;101 it had just one broadband subscriber in December 2015.102 Bharti Airtel fell to second place with 22 percent market share, followed by Vodafone (16 percent), Idea (9 percent) and BSNL (8 percent).103 There are 12 mobile operators,104 with Bharti Airtel controlling almost 24 percent of the market, followed by Vodafone (18 percent), Idea (16 percent) and Reliance Jio (10 percent).105

A universal license framework, for which guidelines were published in November 2014,106 reduced legal and regulatory obstacles by combining mobile phone and ISP licenses. Licensees pay a high one-time entry fee, a performance bank guarantee,107 and annual license fees adjusted for revenue.108

In August 2016, the Cybercafe Association of India (CCAOI) said that 30 percent of venues had closed in the past three years.109 In 2011, the Indian government introduced rules under Section 79 of the IT Act requiring cybercafes to obtain a government-issued ID number in addition to a license, as well as to register and monitor customers.110 Critics said the rules were “poorly framed.”111 Penalties for noncompliance are unclear, and enforcement has reportedly been patchy. Common service centers are exempt, and operate under separate guidelines.112

Regulatory Bodies

Before July 2016, India’s principal ICT institution was the Ministry of Communications and Information Technology. It consisted of two departments – the Department of Electronics and Information Technology (DeitY) and the Department of Telecommunications (DoT).

In July 2016, the Ministry was divided in two. DeitY became the Ministry of Electronics and Information Technology (MeitY), while the DoT and Department of Posts were placed under the Ministry of Communications.113 MeitY formulates policy relating to information technology, electronics, and the internet114 and DoT manages the overall development of the telecommunications sector, licenses internet and mobile service providers, and manages spectrum allocation.115

Internet protocol (IP) addresses are regulated by the Indian Registry for Internet Names and Numbers (IRINN).116 Since 2005, the registry has functioned as an autonomous body within the nonprofit National Internet Exchange of India.117

The Telecom Regulatory Authority of India (TRAI), an independent regulator, was created in 1997 to regulate the telecommunication, broadcast, and cable TV sectors.118 The Telecom Regulatory Authority of India Act (TRAI Act) mandates transparency in the exercise of its operations, which include monitoring licensing terms, compliance, and service quality.119 Its reports are published online, usually preceded by a multi-stakeholder consultation.120 An amendment to the TRAI Act in 2000 established a three-member Telecommunications Dispute Settlement and Appellate Tribunal chaired by a former senior judge.121

There are some reservations about the TRAI’s independence.122 Appointment and salary decisions for members remain in the hands of the central government. The TRAI Act initially barred members who had previously held central or state government office, but 2014 amendments diluted that prohibition, allowing them to join the regulator two years after resigning office, or earlier with government permission. Members may undertake commercial employment, though not with telecom service providers.123 TRAI opinions, however, are generally perceived as free of official influence.124 In 2016, it was involved in framing net neutrality regulations prohibiting discriminatory tariffs for data services.125

B Limits on Content

Content blocking targeting pornography, terrorism, and copyright continued to affect legitimate political and social information during the coverage period, and some content removal by private companies caused controversy. The digital media landscape remained lively and citizens continued to use digital tools to mobilize around important social issues.

Blocking and Filtering

A significant amount of legitimate political and social information was blocked by court or government orders during the reporting period. Since some of those orders are not made public, the exact impact is hard to assess. Entire platforms and services were affected, including the video publishing tool Streamable. Several social media platforms were blocked in an attempt to curtail unrest in Jammu and Kashmir. Separately, courts issued contradictory rulings about copyright blocking, which continues to overreach.

In April 2017, government officials ordered service providers in the Kashmir valley to block 22 social media sites for a month, including Facebook, Twitter and WhatsApp.1 Internet services were separately restricted in the state of Jammu and Kashmir at least 12 times during the coverage period (see Restrictions on Connectivity). The social media order was unusual because it was issued under Section 5 of the Indian Telegraph Act, 1885, raising issues about its legitimacy.2 Section 5 provides for state or central authorities to order “stoppage of transmission … of messages,”3 but blocking orders fall under the IT Act (see below). Many residents circumvented the restriction using virtual private networks (VPNs).4 Some reports said WhatsApp was subject to a separate, temporary restriction on some connections in the state in August 2016.5

Blocking of websites takes place under Section 69A of the Information Technology Act 2008 (IT Act) and a 2009 subordinate legislation called the Information Technology (Procedure and Safeguards for Blocking for Access of Information by Public) Rules (“Blocking Rules”).6 The Blocking Rules empower the central government to direct any agency or intermediary to block access to information when satisfied that it is “necessary or expedient” in the interest of the “sovereignty and integrity of India, defense of India, security of the state, friendly relations with foreign states or public order or for preventing incitement to the commission of any cognizable offence relating to above.”7 Intermediaries failing to comply are punishable with fines and prison terms of up to seven years.8

The Blocking Rules apply to orders issued by government agencies, who must appoint a “nodal officer” to send in requests and demonstrate that they are necessary or expedient under Section 69A.9 These requests are reviewed by a committee, which includes senior representatives of the law, home affairs, and information ministries, and the nodal agency for cybersecurity, the Indian Computer Emergency Response Team (CERT-IN).10 The “designated officer,” who chairs the committee, issues approved orders to service providers; the committee must also notify the source or intermediary hosting the content, who may respond to defend it within 48 hours.11 In emergencies, the Secretary of MEITY may issue blocking orders directly under written instruction from the designated officer, but the content must be unblocked if the review committee does not approve them within 48 hours.12

Indian courts can order content blocks without government approval. The designated officer is required to implement the court order after submitting it to the Secretary of MeitY. Court orders can be challenged in a higher court, but internet users are not consistently notified of their implementation.13

ISPs are not legally required to inform the public of blocks and the Blocking Rules mandate that executive blocking orders be kept confidential.14 In the landmark Shreya Singhal case decided by the Supreme Court in 2015, the petitioners challenged the constitutionality of Section 69A citing opaque procedures among other issues.15 The Supreme Court upheld Section 69A and the Blocking Rules,16 saying safeguards were adequate, narrowly constructed, and constitutional.17 But the court read the Blocking Rules to include both the right to be heard and the right to appeal. Blocking orders must now provide a written explanation, allowing them to be challenged by writ petition, and allow for reasonable efforts to contact the originator of the content for a pre-decisional hearing.18 However, the rules continue to require that the orders and actions based on them be kept confidential;19 it is difficult to know the extent of compliance with the judgment.

In October 2016, the government reported blocking a total of 1377 URLs on social media sites since 2013 under Section 69A20 and 1670 social media URLs in compliance with court orders.21 In most cases, there is no information about the content targeted through these orders.

However, at least some blocks issued in the past year were disproportionate. In January 2017, for example, several ISPs blocked Streamable, an online service used by publishers and bloggers to embed video on their own sites.22 One block page cited directions received from the Department of Telecommunications, but the reason was unclear. The restriction did not affect specific content, but would prevent anyone from using the service to share video within India, cutting off a potential revenue source for content professionals.

Judges sought to improve the framework for blocking content under copyright injunctions in 2016, but broad restrictions continued to be observed in 2017. Since 2011, courts have blocked content relating to copyright violations through broad John Doe orders, which can be issued preemptively and do not name a defendant.23 ISPs have occasionally implemented such orders by blocking entire websites instead of individual URLs, irrespective of whether the websites were hosting pirated material.24 The judiciary has noted that John Doe orders can lead to overblocking,25 and activists have called for greater transparency.26

In July 2016, a ruling by the Bombay High Court laid down rules for seeking John Doe orders, limiting blocks to URLs, not entire domains, and allowing all affected content to be unblocked after 21 days if a court order is not obtained.27 The Court also dictated an unambiguous block message and suggested the appointment of an independent ombudsman to oversee implementation.28 Observers hailed this as a progressive and nuanced approach.29

http://bit.ly/2h3JvWm, Eros International and Another Vs BSNL & Others, Notice of Motion (L) No. 2147 Of 2016 in Suit (L) No. 751 OF 2016 avialble at: https://spicyip.com/wp-content/uploads/2016/09/Bom-HC-order-in-Dishoom_…, Balaji Subramanian, “Bombay HC pulls no punches, orders ISP to comply with John Doe,” SpicyIP, August 10, 2016, http://bit.ly/2hw1PnX, Swapnil Mathur, “Kickass torrent storm: In fight against global piracy, India sets the right examples,” Indian Express, August 3, 2016, http://bit.ly/2aTmg98, Kian Ganz, “The messy battle against online piracy,” LiveMint, August 02, 2016, http://www.livemint.com/Consumer/YtbRN9fv6ZgZCZOexcsWMI/The-messy-battl…, Salman SH, “ISPs block torrents via a John Doe order; lapses in order compliance,” Medianama, August 22, 2016,

http://www.medianama.com/2016/08/223-isps-block-torrent-sites/. But the ruling did not resolve the issue.30 The same month, the Delhi High Court separately ruled that John Doe orders could continue to be used to block websites completely if more than one page on the site was identified as a potential source of copyright violations. Seventy-three websites were involved in that judgment, not because they were proven to violate copyright, but because the plaintiff had pre-emptively identified them as possible violators of exclusive broadcast rights to cricket matches dating from 2014.31 Other John Doe orders continued to be issued,32

http://bit.ly/2bC0DNT; Swati Deshpande, “Madras HC orders blocking of 830 sites to prevent online piracy of 'A Flying Jatt',” TOI, August 24, 2016,

http://bit.ly/2zWXHEo. and legitimate content continued to be affected, including the entire Internet Archive, a nonprofit digital library that was blocked in August 2017, apparently under a John Doe order.33

The IT Act and the Indian Penal Code prohibit the production and transmission of “obscene material,”34 but there is no specific law against viewing pornography in India, except child pornography, which is prohibited under the IT Act (see Legal Environment).35 Extreme child sexual abuse is blocked based on guidance from INTERPOL,36 but other content restrictions threaten content that has not been found to break the law. The government ordered blocks on more than 220 websites advertising escort services during the reporting period,37 allegedly for promoting prostitution, though it was not clear that any of the sites were under criminal investigation, and officials acknowledged that blocking sites was not an effective way to “solve the problem once and for all.”38 In the case of Kamlesh Vaswani v. Union of India, the petitioner asked the Supreme Court to direct the government to block all online pornography.39 A judgement had not been issued in mid-2017, but the government has informed the Supreme Court that it is not feasible to block pornography entirely and that doing so would violate the constitution.40

Content Removal

Improvements to the framework for intermediary liability mean that less political and social content is subject to removal than in the past. However, takedowns and private censorship by companies still caused concern during the reporting period.

An interim order by the Supreme Court had implications for content removal by private companies. In late 2016, the Court ordered search engines operated by Google, Microsoft and Yahoo to “auto-block” advertisements offering services to determine the sex of a child before birth, which contravene a law passed in 1994 in an attempt to stop female feticide. The ruling went further than delisting specific content, asking search engines to block results for specific search terms, and ordering the creation of a nodal agency to oversee the process.41 Critics fear the rulings would restrict related information and breach the Shreya Singhal judgment.42

A 2008 IT Act amendment protected technology companies from legal liability for content posted to their platforms by others, with reasonable exceptions to prevent criminal acts or privacy violations.43 Intermediaries Guidelines issued in 2011 under Section 79 of the IT Act required intermediaries to remove access to certain content within 36 hours of a user complaint.44 The government later clarified this rule.45 In Shreya Singhal v. Union of India, the Supreme Court read down Section 79 and the intermediary guidelines, and companies are no longer required to act on user complaints. Court and government takedown orders, furthermore, are only legitimate if they fall within the reasonable restrictions provided for under Article 19(2) of the constitution. Unlawful content beyond the ambit of Article 19(2) cannot be restricted.46

Based on the ruling, Facebook said it would require more formal notifications to restrict content.47 It restricted 719 items between July and December 2016, citing legal requests from the central government and local law enforcement agencies,48 down from 14,971 items during the equivalent period in 2015.49

Content removal based on alleged violations of Facebook’s community standards still attracted controversy, however, particularly content posted amid protests surrounding the death of a militant in the Kashmir region (see Restrictions on Connectivity).50 In July 2016, Jajeer Talkies, a popular Kashmir-based page that publishes satirical content, was temporarily disabled. Three page administrators also had their profiles disabled, making it harder for them to appeal the action.51 A news video published by a local daily was also removed. Facebook said it removes content about terrorism that does not clearly condemn terrorist organizations or their activities,52 but several academics and journalists were among those temporarily suspended from posting after sharing information about the ongoing crisis.53

Other international companies reported receiving a high number of requests to remove content from Indian courts or government representatives. Google reported receiving 243 content removal requests affecting 543 items between January and June 2016, and said it complied with 35 percent of requests based on court orders and 11 percent from government agencies and law enforcement. The reason most commonly cited for the request was defamation.54 Twitter received 97 requests for content removal from July to December 2016, of which 1 was court ordered and 96 were from police or government agencies, but said it did not comply.55

News reports published in 2017 said that some online video companies were selectively blurring or removing content for Indian audiences. Most examples were explicit, but a scene depicting a cow carcass was also deleted before one show was aired on Amazon Prime.56 Cows have particular status in Hindu mythology (see Digital Activism). There is no legal requirement to remove this content, since the Cinematograph Act, which regulates broadcasts in movie theaters and on television, does not apply online,57 and the Ministry of Information and Broadcasting has explicitly said it has no plans to censor online media.58 Yet companies such as Netflix, Amazon Prime, Google Play and Apple have streamed censored versions of films within India,59 while hosting uncensored versions elsewhere.60 Company policies supporting this practice are unclear and implementation has been irregular, apparently targeting content that is more likely to appeal to Indian audiences instead of applying consistent standards across the board. Amazon said that it censored content that it perceived to be “culturally sensitive” on Prime.61

Intermediaries can separately be held liable for infringing the Copyright Act 1957,62 under the law and licensing agreements.63 The Shreya Singhal decision has had no impact on the legal framework on intermediary liability for copyright infringement. A 2012 amendment limited liability for intermediaries such as search engines that link users to material copied illegally, but mandated that they disable public access for 21 days within 36 hours of receiving written notice from the copyright holder, pending a court order to remove the link.64 Rules clarifying the amendment in 2013 gave intermediaries power to assess the legitimacy of the notice from the copyright holder and refuse to comply.65 However, critics said the language was vague.66

Media, Diversity and Content Manipulation

Online media content is diverse and lively. The internet has given a voice to people in remote areas, helping them become a part of the public discourse. The Delhi-based company Gram Vaani operates a Mobile Vaani initiative, using an interactive voice response (IVR) system to disseminate reports by mobile phone users to different audiences and stakeholders. It enables over 80,000 households across 12 states to create their own media.67 Some citizens have also turned to digital tools to escape the partisan traditional news media environment. During the reporting period, residents of Tamil Nadu found ways to counter news outlets controlled by major political parties by sharing satire on Facebook and WhatsApp.68

In general, self-censorship is not widespread. Internet users in conflict regions may sometimes avoid addressing sensitive political or religious issues but other journalists and activists report freely. During the ongoing conflict in Kashmir Valley, traditional news outlets even chose to go entirely online in order to cut costs and maintain jobs, a development which heightens the potential threat to freedom of information posed by the state’s frequent internet shutdowns (see Restrictions on Connectivity).69

Demonetization of currency notes in small denominations in November 2016 decreased cash flow across the economy (see Availability and Ease of Access). Analysts said the move had a positive impact on information surrounding elections in the state of Goa by significantly reducing the use of cash bribes for positive news coverage. Instead, candidates campaigned vociferously on social media, helping newer parties without significant financial support.70 The same policy had a less positive effect on internet freedom in Indore, a city of 2 million people in Madhya Pradesh. In an unprecedented move, the District Magistrate issued an order under Section 144 of the Code of Criminal Procedure banning any criticism of the demonetization policy on social media considered to be “objectionable” or cause “incitement.”71 There was no information available about the impact of this order, but at least two residents of Madhya Pradesh were detained in relation to digital speech about demonetization (see Prosecutions and Detentions for Online Activity).

Politicians in India have embraced social media, often announcing major policies directly on platforms like Twitter.72 The prime minister is the most followed politician on Twitter in the world after U.S. President Donald Trump,73 and has developed his own application, NMApp, to communicate with followers.74 Critics say this helps him avoid engaging directly with journalists who might challenge controversial policies.75

Government employees are encouraged to use social media, but the prime minister warned them not to use social media to promote themselves.76 A proposal to amend a public service code of conduct to include the online behavior of government employees was floated in July 2016.77

While the internet serves as a tool of empowerment for many Indians, however, some trends caused concern in the past year, including allegations of politicized content manipulation. Aggressive online commentators who self-identify as Hindu nationalists routinely abuse their opponents; however, research published during the reporting period said employees of the ruling BJP party orchestrated some of the activity, including threats and abuse targeting women and journalists, in order to create a hostile online environment for people criticizing the government and its leaders.78

BJP politicians have been accused of paying specialized companies to artificially boost their popularity on social media since they were in the opposition in 2013,79 and some party representatives were reported to have paid citizens to post messages of support before their successful 2014 election.80 Rival parties have likewise been accused of secretly sponsoring online support.81 While trolling that appears to align with the BJP agenda has continued under the BJP administration, there is no evidence that government actors are directly involved. Rather, officials’ tacit support of online abuse—evidenced, for example, by the prime minister following known troll accounts on Twitter—contribute to a climate where people who are perceived to oppose popular discourse face intimidation, even while robust political debate continues in many online forums. Online harassment remained widespread during the reporting period (see Digital Activism and Intimidation and Violence).

Initiatives to monitor social media use are periodically reported, and some in the past year involved incidents of violence shared using live video streaming tools.82 The government accused militants of using Facebook Live to instigate violent anti-state activity in the Kashmir Valley,83 and was formulating a social media policy to monitor antinational propaganda and curb malicious rumors in 2017 with particular reference to the situation in Jammu and Kashmir.84 Some violent and disturbing events were certainly streamed in that region and beyond. A student in Mumbai took his own life while broadcasting on Facebook.85 At the same time, however, protestors also used social media to report on human rights violations by security forces in Kashmir.86 Facebook’s own attempts to limit terrorist content were subject to criticism for failing to adequately distinguish between those who supported violence and those who simply shared information about it (see Content Removal).87

Digital Activism

Digital activism is popular and has resulted in some proven successes on the national scale in the past. Through various campaigns and groups used social media in the reporting period there were no widespread effect.

Some groups successfully leveraged social media to draw national attention to local protests during the reporting period. In August 2016, for example, a video documenting abuse of Dalits in a district of Gujarat launched major protests against discrimination. Dalits are marginalized in the traditional Hindu caste system. With the rise of nationalist politics, vigilante groups characterizing themselves as gau rakshaks (cow protectors) have attacked Dalits and other minorities that consume or handle beef.88 The assailants circulated a video of an attack targeting eight members of a Dalit community who skin cattle for a living as a warning, but protesters used it to ensure national media covered the incident, and subsequently organized demonstrations and a strike using social media and communication apps.89 However, few changes resulted.90

C Violations of User Rights

Several arrests for online speech were reported during the coverage period, including for content distributed on WhatsApp and Facebook. Attacks and social sanctions were also reported in reprisal for online information and commentary. In a positive development, the Supreme Court issued a landmark privacy ruling.

Legal Environment

The Constitution of India grants citizens the fundamental right to freedom of speech and expression,1 including the right to gather information and exchange thoughts within and outside India.2 Press freedom has been read into the freedom of speech and expression.3 These freedoms are subject to certain restrictions in the interests of state security, friendly relations with foreign states, public order, decency and morality, contempt of court, defamation, incitement to an offense, and the sovereignty and integrity of India. However, these restrictions may only be imposed under a law, not by executive action.4 The right to privacy has been read into the right to life guaranteed by Article 21 of the constitution.5 The Supreme Court in a landmark ruling issued outside the coverage period of this report recognized privacy as fundamental right which is protected as intrinsic part of other rights including right to life, liberty and freedom of expression (see Surveillance, Privacy, and Anonymity).6

The Indian Penal Code (IPC) criminalizes several kinds of speech, and applies to online content. Individuals could be punished with between two and seven years in prison for speech that is found to be seditious,7 obscene,8 defamatory,9 “promoting enmity between different groups on ground of religion, race, place of birth, residence, language,”10 committing acts “prejudicial to maintenance of harmony,”11 or consisting of statements, rumors, or reports that may cause fear, alarm, disturb public tranquility, or promote enmity or ill will.12 Internet users are also subject to criminal punishment under the Official Secrets Act for wrongful communication of information that may have an adverse effect on the sovereignty and integrity of India.13

The IT Act criminalizes certain online activity such as the creation, transmission or browsing of child pornography.14 Section 67 bans the publication or transmission of obscene or sexually explicit content in electronic form, and Section 66D punishes the use of computer resources to impersonate someone else to commit fraud.

Section 66A, a particularly problematic provision, was struck down by the Supreme Court in 2015. The provision criminalized information causing "annoyance,” “inconvenience,” or “danger,” among other ill-defined categories and lead to several arrests for social media posts from 2012 through early 2015. The court in the Shreya Singhal judgment15 affirmed that freedom of speech online is equal to freedom of speech offline, and held that Section 66A went beyond reasonable restrictions on that freedom specified in Article 19(2) of the constitution.16 Outstanding prosecutions under the section were dropped,17 but similar complaints continue to be registered under Sections 67, 66D, or the IPC (see Prosecutions and Detentions for Online Activities).

A 2016 Supreme Court judgment upheld laws criminalizing defamation (Sections 499 and 500 of the IPC and Section 119 of the Code of Criminal Procedure) as consistent with the Indian Constitution.18 The sections have been used against online speech in the past.19

Prosecutions and Detentions for Online Activities

More than 20 criminal complaints involving online content were filed during this coverage period, and over a dozen people were detained for sharing content that was considered political or religious actors considered insulting or misrepresentative. This continued the trend described in 2016; before that, the number of detentions fell off slightly following the Supreme Court’s Shreya Singhal ruling on the IT Act (see Legal Environment). However, no convictions were documented in the reporting period.

In one notable case, Taufiq Ahmed Bhatt, a Kashmiri student in Bhilai, Chhattisgarh was arrested on charges of sedition in August 2016, for sharing and responding to “anti-India” posts on social media, including an image representing India as a mouse being swept away by a broom.20 He was in custody until at least October 2016.21 No information was available regarding his case in mid-2017.

In November 2016, Justice Rajiv Sahai Endlaw of the Delhi High Court held that social media group administrators cannot be held liable for content shared by group members,22 an issue that lead to a number of arrests last year.23 In April 2017, however, a District Magistrate and a local police chief in Uttar Pradesh jointly warned that group administrators on social media or messaging apps would be subject to criminal complaints if inaccurate content was shared within the group,24 causing confusion on the issue of liability.25 Officials in Bhagalpur district in Bihar issued a similar warning.26 In May 2017, police in Karnataka detained Krishna Sanna Thamma Naik, the administrator of a WhatsApp group, after a member of a group shared an image of Prime Minister Narendra Modi that news reports described as “obscene.” The member responsible was also detained; both were released on bail.27

Several other cases prompted detentions, including the following examples:

  • In September 2016, 32-year-old Tarak Biswas from Kolkata was detained under several Sections of the IT Act for posting objectionable content and insulting religious feelings on Facebook. Biswas describes himself as an atheist and frequently criticizes religion.28
  • In October 2016, police in Karnataka detained three men for posting a video that allegedly depicted residents and leaders from neighboring Tamil Nadu in a bad light. The states were involved in a dispute over the distribution of water from a local river.29
  • In November 2016, police in Madhya Pradesh detained 19-year-old student Abhishek Mishra under Section 469 of the IPC and Section 66C of the IT Act, which pertain to forgery and identity theft, for posting about a demonetization policy intended to combat corruption (see Media, Diversity, and Content Manipulation). The post included a photo of police seizing money from a local BJP leader, but police alleged that Mishra had deliberately misidentified the leader as the Madhya Pradesh chief minister.30 In a separate case the same month, 25-year-old BJP party worker Aslam Khan, who expressed dissatisfaction with the demonetization policy, was detained under Section 505(2) of the IPC in the same state’s Morena district. He had allegedly posted a photograph of Prime Minister Narendra Modi in a WhatsApp group that was considered insulting because it had been altered to show him wearing a garland of shoes.31
  • In March 2017, at least five people were detained in different parts of Uttar Pradesh for posting “objectionable” content about newly-appointed Chief Minister Yogi Adityanath on social media.32 The reports did not elaborate on the nature of the content involved.
  • In April 2017, police in Hyderabad detained satirist Inturi Ravikiran for making derogatory remarks about the Andhra Pradesh Assembly and its members.33

Surveillance, Privacy, and Anonymity

An important Supreme Court ruling, issued after the reporting period, recognized privacy as a fundamental right.34 A draft privacy bill had been pending for many years,35 and the government recently constituted a committee to frame a data protection framework.36

The 2017 Supreme Court ruling came in the context of the Aadhaar scheme, a unique identification project creating a database of citizens’ biometric and other data. The government has required Aadhaar enrollment for the provision of multiple public services.37 The scheme raised concerns regarding data privacy, security, and usage.38 In 2017, it was reported that millions of Aadhaar records have been treated as publicly shareable data by different government departments.39 A national government-administered rural employment scheme was among several initiatives or agencies reported to have accidentally revealed Aadhaar numbers.40

There is limited opportunity for anonymity on the internet in India. Prepaid and postpaid mobile customers have their identification verified before connections are activated.41 There is a legal requirement to submit identification at cybercafes,42 and while subscribing to internet connections. The effective implementation of privacy rights remains a significant issue. Communications surveillance may be conducted under the Telegraph Act,43 as well as the IT Act,44 to protect defense, national security, sovereignty, friendly relations with foreign states, public order, and to prevent incitement to a cognizable offense. Section 69 of the IT Act appears to add another broad category, allowing surveillance for “the investigation of any offence.”45

The home secretary at the central or state level issues interception orders based on procedural safeguards established by the Supreme Court and rules under the Telegraph Act.46 These are reviewed by a committee of government officials of a certain rank, and carried out by intermediaries.47 A similar framework applies to the IT Act.48 Interception orders, which are not reviewed by a court, are limited to 60 days, renewable for up to 180 days.49 In emergencies, phone tapping may take place for up to 72 hours without clearance; records must be destroyed if the home secretary subsequently denies permission.50

Eight separate intelligence bodies are authorized to issue surveillance orders to service providers under these circumstances.51 Around 7,500 to 9,000 telephone interception orders are issued by the central government alone each month, according to a 2014 report citing information revealed in a right to information request.52

Online intermediaries are required by law to “intercept, monitor, or decrypt” or otherwise provide user information to officials.53 The Telegraph Act levies civil penalties or license revocation for noncompliance54 and the IT Act carries a possible seven-year jail term.55 Unlawful interception is punishable by just three years’ imprisonment.56

Some improvements to the framework have been made. On January 2, 2014, the government issued “Standard Operating Procedures (SOP) for Lawful Interception and Monitoring of Telecom Service Providers,” which were viewed by journalists but not publicly available.57 The procedures restricted interception to a service provider’s “chief nodal officer,” and mandated that interception orders be in writing.58 Rules issued in 2011 under the IT Act increased protection of personal data handled by companies.59 However, they do not apply to the government; critics say they create a burden on multinational companies, particularly in the context of the outsourcing industry.60

These improvements failed to address the framework’s inconsistencies. In 2012, a government-appointed group of experts said the Telegraph and the IT Acts are inconsistent with regard to “permitted grounds,” “type of interception,” “granularity of information that can be intercepted,” the degree of assistance from service providers, and the “destruction and retention” of “intercepted material.” These differences, it concluded, “have created an unclear regulatory regime that is non-transparent, prone to misuse, and that does not provide remedy for aggrieved individuals.”61

License agreements require service providers to guarantee the designated security agency or licensor remote access to information for monitoring;62 ensure that their equipment contains necessary software and hardware for centralized interception and monitoring; and provide the geographical location, such as the nearest Base Transceiver Station, of any subscriber at a given point in time.63 Under a 2011 Equipment Security Agreement that did not appear on the DoT website, telecom operators were separately told to develop the capacity to pinpoint any customer’s physical location within 50 meters.64 “Customers specified by security agencies” were prioritized for location monitoring, with “all customers, irrespective of whether they are the subject of legal intercept or not,” to be monitored by June 2014.65 The agreement remains effective, though various GSM operators lobbied for the clause to be removed from the license agreement because of compliance issues.66 In 2014, an amendment to licensing conditions mandated government testing for all telecom equipment prior to use, effective in 2015.67 Cybercafe owners are required to photograph their customers, arrange computer screens in plain sight, keep copies of client IDs and their browsing histories for one year, and forward this data to the government each month.68

ISPs setting up cable landing stations are required to install infrastructure for surveillance and keyword scanning of all traffic passing through each gateway.69 The ISP license bars internet providers from deploying bulk encryption; restricts the level of encryption for individuals, groups or organizations to a key length of 40 bits;70 and mandates prior approval from the DoT or a designated officer to install encryption equipment.71

The government also seeks user information from international web-based platforms. Google reported that the government made 3,449 user data requests and 6,393 requests to access accounts between July and December 2016. Google made disclosures in 57 percent of the cases.72 The government requested access to 8,221 Facebook accounts between January and June 2016 and data was produced by Facebook in 53 percent of cases.73 The government made 168 account information requests to Twitter between June and December 2016, the highest by any government so far; Twitter said it produced data in 23 percent of cases.74

Besides retrieving data from intermediaries, the government’s own surveillance equipment is becoming more sophisticated. The Central Monitoring System (CMS) allows government agencies to intercept any online activities directly, including phone calls, text messages, and VoIP communication, using Lawful Intercept and Monitoring (LIM) systems on intermediary premises.75 In May 2016, the Minister for Communications and IT stated that the monitoring centers were already operational in Delhi and Mumbai.76 More centers were due to be rolled out across the country, but no updates were available in mid-2017.

MeitY officials indicated that security agencies could access messaging services such as WhatsApp in 2017, though they are unable to view encrypted content. In response to a question in the Lower House of Parliament, the Minister of State for Information Technology stated that “security agencies are able to intercept these encrypted communication services through the lawful interception facilities provided by the Telecom Service Providers, but they are not able to decrypt some of encrypted intercepted communication to readable format.”77

Law enforcement agencies may proactively monitor social media for signs of wrongdoing, although the legal grounds for doing so is unclear. In March 2017, the Minister of State for Electronics and Information Technology said that “social Networking sites hosted anywhere in the world are monitored by the law enforcement agencies.”78

Intimidation and Violence

Internet users were attacked and subject to social sanctions in reprisal for online speech during the coverage period. In September 2016, a 22-year-old student in Bengaluru was assaulted for social media posts in which he criticized Kannada celebrities involved in a water dispute between Tamil Nadu and Karnataka; another social media user had been separately arrested for offending the Tamil Nadu side of the dispute (see Prosecutions and Detentions for Online Activities).79

In May 2017, Sunil Waghmar, a Dalit professor, was beaten by a mob for forwarding a message that appeared to make light of a religious festival to a WhatsApp group of fellow professors. He was suspended from his job for misconduct, and police arrested him for the same message while he was still badly injured.80

Women, journalists, and political activists report frequent trolling and violent threats in response to their online posts.81 In July 2016, Minister of Women and Child Development, Maneka Gandhi, invited women facing harassment to report it to her directly by email after a journalist reported that one abusive post appeared to come from a popular singer.82 The government is also developing an app in response to the problem,83 but women continued to be silenced as a result of the behavior in the past year. Gurmehar Kaur, a 20-year-old student of Lady Sri Ram College, New Delhi, was subject to threats of rape and murder after she criticized a hardline right-wing student group in an online video in February 2017.84 She withdrew from a related protest campaign as a result of the harassment.85

One sample survey of 100 women published during the coverage period found that most had faced severe online harassment but were hesitant to report them, even if they knew it was an option.86 A separate sample survey of 500 respondents made up of 97 percent women found that nearly 60 percent had faced online harassment, sometimes including violent threats.87

Technical Attacks

Indian Computer Emergency Response Team (CERT-In) reported a total of 3,347 websites hacks in 2016.88 Almost 200 central and state government sites were affected, including the site of Union Ministry of Home Affairs.89 CERT-In issues periodic advisories, and the government updates a Crisis Management Plan for central and state governments to respond to cybercrime on an annual basis.90 However, attacks to suppress online speech are not known to be widespread; most had economic motives, according to the National Crime Records Bureau.91

A hacker group calling itself Legion hacked into email and twitter accounts operated by high profile Indians, including two journalists, in December 2016, but without an obvious political agenda.92 A Congress Party leader and a liquor baron were also targeted, and the hacker claimed to have email addresses and passwords for more than 74,000 chartered accountants.93 Citizens also had their personal data exposed following technical attacks in the past year, including one reportedly involving 3.2 million ATM debit cards issued by major banks.94

On India

See all data, scores & information on this country or territory.

See More

Country Facts

  • Global Freedom Score

    66 100 partly free

  • Internet Freedom Score

    51 100 partly free