Alert Sign Dear reader, online ads enable us to deliver the journalism you value. Please support us by taking a moment to turn off Adblock on Dawn.com.

Alert Sign Dear reader, please upgrade to the latest version of IE to have a better reading experience

.

Email

Special report: The mechanics of silencing online dissent

Online campaigns are not looked upon kindly when they go against the dominant state narrative, even if they are organic.
Updated May 19, 2019 09:01pm

Journalists in Pakistan who dissent from the official narrative are feeling the brunt not only of the state and its bad laws. A more insidious form of censorship is being used to silence them online.


Each morning, in the last few months, journalist Shahzeb Jillani would wake up to a barrage of notifications on Twitter.

Jillani, who has a following of over 14,000 users on the micro-blogging site, was usually not alarmed by the scurrilous trolling and abuse.

But on March 24, he admits, “something changed.” That was when he posted a tweet criticising the decision to decorate a senior military intelligence officer “widely accused of political engineering” during the July 2018 election.

“I could feel the pressure building up. The tone [of trolls] had become extensively threatening and specific,” he tells Eos during an interview.

Among the 197 replies to Jillani’s tweet is a swarm of suspicious accounts hurling terms akin to ‘traitor’ and questioning his loyalty to the country.

It was the attention to detail that Jillani found unsettling.

According to him, a majority of the accounts engaging with him after that tweet had a profile picture of the Pakistan flag, a numeric Twitter handle and between 20 to 25 followers on an average.

Referring to one of these tweets, he recalls: “One of the comments tagged DG ISPR [Director General Inter Services Public Relations] suggesting to him that if they can take care of other TV journalists by getting them fired from their jobs, why can’t they take care of me?”

Another troll warned him that he would soon lose his job at Dunya News channel.

A few weeks later, Jillani was, in fact, sacked from his job.

On April 8, while on his way back from a reporting assignment in Thar, Jillani’s phone buzzed incessantly with an onslaught of message alerts.

Ironically, it was through Twitter that he learned that the Federal Investigation Agency (FIA) had filed a case against him on charges of cyber-terrorism, which carries a maximum jail sentence of 14 years.

According to the First Investigation Report, Jillani was accused by a viewer of “articulating defamatory remarks against the respected institutions of Pakistan”, a crime under the Pakistan Electronic Crimes Act (Peca) 2016.

The case is based on the complaint by a self-described “loyal citizen”, who has also filed various petitions challenging civilian political leaders in the past.

The complainant alleged that Jillani was toeing the “line of foreign agencies on social media and had also been blogging in order to implement his agenda against the sovereignty of Pakistan”.

The complainant further claimed that the act of the ‘suspect’ was “in collusion with the enemy countries that seems to intense [sic] conspiracy, which directly or indirectly damaged and defamed the State of Pakistan by casting audacious aspersions against sacred institutions.”

The statement of the FIA cybercrime reporting centre in Karachi mentions that, during the course of the inquiry, it was found that “Jillani had — with criminal intent and ulterior motives and without any lawful justification — made sarcastic, derogatory, disrespectful remarks and used defamatory language against Pakistan, the Election Commission of Pakistan, and the armed force [sic].”

The charge sheet includes printouts of screenshots of specific tweets posted by Jillani.

In his statement to the FIA, Jillani maintained that his duty as a journalist demands that “wherever I become aware of any wrongdoing on part of any public institution or public functionary, I must point it out.

“One may agree or disagree with [the] criticism but not [with] my constitutional right to criticise. No institution or individual involved in politics can claim to be above criticism,” he stated.

However, according to the complainant, Jillani’s actions “are tantamount to creating a sense of fear, panic, insecurity in government institutions, [the] general public and society”.

Consequently, he contended, besides cyber-terrorism, the investigative reporter — who has worked previously for the BBC and Deutsche Welle — be charged for hate speech and offence against “dignity of a natural person.”

“The purpose of these conspiracy-themed accounts is to drown out critical voices. They wanted Jillani to go offline, and he did,” Jillani says dejectedly.

Since the FIA case, the journalist has been advised to refrain from social media activity.

In the latest update till the filing of this report, the FIA has withdrawn the cognisable charges (cyber-terrorism) as well as the charge of hate speech in its final investigation report submitted in the court.

However, the case is still under trial for charges of defamation under Section 20 of Peca.

Read more: Case against journalist Shahzeb Jillani quashed due to lack of evidence

According to Farieha Aziz, co-founder of digital rights group Bolo Bhi, various irregularities go unchecked under the law.

“The FIR against Shahzeb is exactly what we said Peca would be used for,” says Aziz.

“From arbitrary summons and inquiries, we’ve seen people slapped with charges of cyber-terrorism, hate speech and, in particular, criminal defamation — i.e. Section 20, [which pertains to the] dignity of a natural person.”

Exacerbating misuse of the law are the vaguely phrased provisions of Peca. “If you look at the language of the law, Section 20 mentions “aggrieved person”.

The complainant must be an individual. An institution cannot be an aggrieved person and an officer or another cannot, on its behalf, invoke this section of the law,” she points out.

“Then, if you look at the FIRs in other [Peca] cases, the FIA has been adding cognisable sections of the Pakistan Penal Code [PPC] to acquire power to arrest — and, in some cases, also ATA [the Anti-Terrorism Act]. They require warrants to search and seize devices — however, those summoned for inquiries and detained have reported their devices have been seized while at the station,” she points out.

For Jillani and those accused under the law, regardless of the merits of the cases, the hearings go on.

Such harassment of dissidents and journalists who don’t fall in line is nothing new. But what is new are the tools being employed to enforce silence.

Aside from the use of poorly drafted new technology laws, there are also more insidious ways, which involve online bullying and smearing through targeted campaigns on social media.

Anatomy of a campaign

Pakistan’s Twitter landscape has a comparatively dense share of users, and the influence of propaganda accounts and coordinated networks on national discourse has intensified recently.

For troll groups aligned with a populist narrative, group chats in apps such as WhatsApp are active all day long, discussing potential targets to abuse, threaten or intimidate.

They scour Twitter feeds of journalists and political activists critical of the state narrative, and celebrate small victories in terms of virality, retweets and disbursement of content.

On an average day on local Twitter, it is normal to come across two to three hashtags or ‘trending’ topics.

Whereas normally hashtags ‘trend’ through popular engagement of users with the topics, in Pakistan these often indicate manufactured campaigns.

The content of these campaigns is often replicated, and a scroll down the hashtag feed almost always leads to a user (ideally with a large following) tweeting the merchant code: “Grab your keyboards and start trending.”

Who are these hashtag merchants, and what motivates them?

There are three types of networks stifling critics on the platform: those who trend hashtags to augment the pro-government narrative; those who actively engage with journalists and activists critical of the government; and those who monitor and select ‘targets’.

Last month, 10 journalists were targeted with coordinated smear campaigns online.

The trends which were listed among top trends in Pakistan — accompanied by abuse, defamatory content and doctored images — specifically focused on journalists questioning the government’s policies.

The coordinated network of pro-government supporters accused them of corruption, of taking bribes, bias and of spreading fake news.

Earlier, in December 2018 journalist Raza Rumi, then editor at The Daily Times, was the target of an online campaign against him after he published an editorial by an Ahmadi author in criticism of an online influencer.

“Between December 31 and January 2, more than 2,000 tweets were sent out against me,” says Rumi.

“This barrage of abuse and defamatory content has been and can be a threat to real life employment. This is preemptive harassment, that often leads to dangerous consequences. It not only encourages self-censorship online but also influences editorial policy.”

According to Saeed Rizwan, a cultural anthropologist specialising in big data, users populate the hashtags using a structured network.

The users — who are predominately alleged supporters of PTI based on their bio description and profile pictures — running the campaigns were only invested in amplifying the abuse hashtags and maximising mentions of the journalists being targeted. They did not engage in individual trolling.

To distinguish between organic and coordinated campaigns, consider Graphs 1 (below left) and 2 (below right).

Visualising the data: (Graph 1, left) An organic discussion on Twitter about the Aurat March; (Graph 2, right) an organised campaign targeting journalist Mansoor Ali Khan. — Courtesy Saeed Rizwan
Visualising the data: (Graph 1, left) An organic discussion on Twitter about the Aurat March; (Graph 2, right) an organised campaign targeting journalist Mansoor Ali Khan. — Courtesy Saeed Rizwan

Graph 1 demonstrates an organic discussion under the hashtag #AuratMarch. The users were participating across the map, and the percentages of reply and reply mentions was significant.

In contrast to that, Graph 2 demonstrates an example of an organised campaign against journalist Mansoor Ali Khan, where the retweet percentage was exceptionally high and reply mentions and ‘quote tweets’ (where a user adds their own comments while quoting someone else’s tweet) were low.

A major flux of retweets were directed to two users — @YasifChaudhry (retweeted more than 900 times) and @FarhanKVirk (retweeted more than 1,500 times).

To further understand the mechanics of a coordinated campaign, consider Graph 3. It shows the most active Twitter users of the smear trends against journalists.

A visualisation of libel campaigns in the last week of April against journalists and activists Umer Cheema, Marvi Sirmed, Mubashir Zaidi, Fakhar Durrani, Arshad Waheed and Saleem Safi. —Courtesy Saeed Rizwan
A visualisation of libel campaigns in the last week of April against journalists and activists Umer Cheema, Marvi Sirmed, Mubashir Zaidi, Fakhar Durrani, Arshad Waheed and Saleem Safi. —Courtesy Saeed Rizwan

The size of each circle indicates the numbers of tweets/retweets an individual sent using the hashtag. User @Talagangi_malik, for instance, sent out a huge chunk of tweets involved in the hashtag.

The circles with the same colour show how each user has his own sub-communities that retweet content generated by the leading account. According to Rizwan, each user sent out at least 40 tweets during the campaigns.

The tail of the arrow shows the sender who sent out the tweet while the head is the receiver. User @JanNazar1, for instance, generated less content (as indicated by a small circle) but he was retweeted more.

“Labour is cheap in Pakistan,” says a social media activist, requesting not to be named.

“It’s easier for these networks to recruit and mobilise because people are willing to run multiple accounts for as low as two dollars an hour. You have a daily quota to fill — say two hashtags for peak hours on a daily basis. Either you are the one tweeting content or you are the one retweeting. Either way, your success is measured by how much traffic you generate,” he explains.

The common targets for mentions, he confesses, are journalists. “They are closer to news. If you get their attention, the word spreads around faster,” he says.

Digital surveillance

There’s a bias in how these negative social media campaigns are dealt with, however.

They are not looked kindly upon when they go against the dominant state narrative, even if they may have emerged organically.

On March 13, in a leaked letter online, the cybercrime wing of the FIA was ordered to initiate inquiries against five journalists and parties for allegedly executing what it called “a targeted social media campaign” against Saudi Crown Prince Mohammad Bin Salman during his February visit to Pakistan.

The ‘campaign’ in question consisted merely of users changing their profile photos to that of journalist Jamal Khashoggi — who was murdered inside the Saudi consulate in Istanbul in October — as a form of protest during the Saudi royal’s visit.

This “conveyed a very disrespectful message” towards the visitor, the letter said, adding that a few social media activists and groups remained particularly active in running the campaign till the very last day of the visit.

“There was no prior communication from the FIA regarding the enquiry,” Murtaza Solangi, one of the journalists named in the list, tells Eos over the phone. “As soon as I tweeted a picture of the letter I received via my sources, a swarm of pro-government accounts started to attack me online.

“They [trolls] began challenging me to put up Khashoggi’s picture again after the letter came out. They feel so empowered that they can intimidate journalists and suppress critical voices,” he says.

According to Solangi, no enquiry has been initiated against the journalists named in the letter as per his knowledge.

Section 37 of the cyber crimes law chalks out restrictions allowing for the PTA to block, remove and/or issue directions to censor online content through an information system if it considers necessary to do so in the interest of “the glory of Islam, or the integrity, security or defence of Pakistan or any part thereof, public order, decency or morality, or in relation to contempt of court.”

Interestingly, unlike the Constitution, it does not mention anything pertaining to content about relations with friendly foreign states.

Nevertheless, says Solangi, “The law is so ambiguous it can be used to harm anybody. It is being used as a tool to harass dissenters.”

According to the Pakistan Internet Landscape Report 2018 published by human rights and advocacy group Bytes for All, over 50% of journalists considered digital surveillance a reason of intimidation for them and of credible life threats to their sources.

According to the same report, major topics that were likely to trigger backlash online included regional narratives that are seen as anti-Pakistan, such as criticism of friendly states — China and Saudi Arabia — or support of India/betterment of Pak-India ties, and general criticism of the military and judiciary related to alleged interference in politics and the erosion of democracy, especially during the 2018 general elections.

“Over the years, social media has evolved from a space where ideas could be expressed freely to a space with massive surveillance by state and non-state actors,” says Nighat Dad of the Digital Rights Foundation (DRF).

According to Dad, at least 55 women from Pakistan who reached out to DRF — including journalists and activists — were recently targeted with a sophisticated ‘phishing attack’.

A phishing attack is an attempt to obtain private information such as passwords of those targeted, usually through camouflaged means.

“An email carrying a malware link is targeting at risk groups,” shares Dad. “The email is sent out from someone impersonating Sheherbano Taseer [former Governor Punjab Salmaan Taseer’s daughter].” Clicking on the link allows email accounts to be compromised.

At a time when the online space is afloat with reports citing surveillance risks, users have more frequently shared screenshots of alerts from Google, warning them against state-backed hackers trying to break into their email accounts.

Academic Nida Kirmani, who had previously received a legal notice from Twitter regarding an official complaint against one of her tweets — allegedly for violating Pakistan’s laws — was the first to post about the alert from Google.

“There’s a chance this is a false alarm, but we detected government-backed attackers trying to steal your password. This happens to less than 0.5% Gmail users,” read the alert.

In response to Kirmani’s tweet, various other journalists and activists shared that they, too, had received similar warnings.

“We see how free speech and activism of certain individuals are under constant threat because their opinions aren’t in line with the mainstream rhetoric,” regrets Dad. “This is increasingly problematic as death threats online have become a norm.”

Earlier this year in February, a list comprising names of over 30 journalists and activists was circulated on WhatsApp.

Former Pakistan ambassador Hussain Haqqani was labelled the “patron in chief” of the group labelled as enemies of the state, while others were prominent journalists, activists, and politicians critical of the military. The origin of the list remains unclear.

The result of all this, however, is that the online space is shrinking for journalists as propagandists have fast developed a culture where critical reporting is dismissed as an expression of bias, fake news or negative coverage.

Abuse going unchecked

In February, the FIA had claimed to have put in place a monitoring system following government directives about a crackdown against those spreading extremism and hatred on social media.

According to a Dawn report, a senior official had confirmed that after the decision of the federal government, all three FIA cybercrime centres in Sindh — Karachi, Hyderabad and Sukkur — had been officially directed to keep a close eye on social media and all digital platforms.

However, the increase in concerted vilification campaigns propagated by a network of fake accounts — against journalists, news organisations as well as individuals critical of the government — continues unchecked by the authorities.

“In Sindh, there are thousands of investigation officers for seven districts but for the cybercrime unit, there are barely 15 of us, our hands are full,” says Faizullah Korejo, who heads the agency’s cybercrime wing in Sindh. “We only operate on orders from higher ups,” he concedes.

According to the Prime Minister’s Focal Person on Digital Media Arslan Khalid, the FIA’s failure is the failure of the interior ministry.

“But journalists, too, should act responsibly. Criticising those spreading fake news is not wrong. Everyone should be held accountable,” he says regarding the trend of smear campaigns against journalists.

Given that the network analysis suggests that a majority of the abusive campaigns were being run by ruling party supporters, Khalid — who was the PTI’s social media secretary before assuming office and also follows major users involved in the trends on Twitter — maintains that there is zero tolerance for abuse at any level and that it is against PTI’s code of conduct.

“Abuse is wrong, but criticism is justified,” he says, adding that the government is actively working with Facebook to curb the spread of hate speech and extremism.

For its part, Facebook removed 103 pages in April for engaging in “coordinated inauthentic behaviour” on Facebook and Instagram as part of a network that originated in Pakistan.

The social media giant said it had removed the pages, accounts and groups set up by the networks “for violating Facebook’s policies on coordinated inauthentic behaviour or spam.”

Nathaniel Gleicher, the company’s head of Cybersecurity Policy on the investigation had shared that they were on board with the Prime Minister’s Office and the “social media adviser” prior to the suspension of accounts.

“We recently sent Facebook a list of accounts spreading hate content and they have suspended the accounts. Twitter does not comply as swiftly. Mostly, because the abusive content is in Urdu and they don’t pick up on it easily,” says Khalid.

As per the recent biannual transparency report of the micro-blogging site, the government sent requests for removal of 193 accounts and reported 2,349 profiles to Twitter between July and December.

While Twitter did not completely remove any account, it removed some content from 204 accounts for violating Twitter’s Terms of Reference as compared to 141 in Jan-Jun 2018.

For propagandists and coordinated groups, however, Twitter has sharply escalated its battle against fake and suspicious accounts.

During his research mapping the online space in Pakistan, Rizwan recorded that at least 50 accounts were suspended by Twitter between March and April.

“On average, 100 Twitter accounts from Pakistan are suspended every month,” he reveals. A survey of the accounts suspended shows that the users were seemingly suspicious in their activity with common keywords for handles including ‘Pakistan’, ‘PTI’, ‘army’, and ‘zindabad’.

But a crackdown on fake accounts does not alone solve the problem. If one account is blocked by too many users, the trolls simply close it and open a new one.

Given the concurrent trend of criticism online and the government’s lack of acknowledgment of the same, information practitioners wonder if the real motivation is to peddle the larger narrative of regulating digital media.

The writer is a member of staff

Header illustration by Hufsa Chaudhry

Published in Dawn, EOS, May 19th, 2019