At 8:45 pm, five gunmen stormed the Holey Artisan Bakery, an upscale establishment in Dhaka, the capital of Bangladesh. The siege lasted more than 12 hours. By the time it was over, 22 people were dead, most of them foreigners. While other terrorist attacks have occurred in Bangladesh since that fateful evening on July 1, 2016, the massacre at Holey Artisan is considered the deadliest terrorist attack in the country’s history. It also triggered new thinking about how to counter violent extremism and gave rise to efforts that can serve as a model for other countries tackling similar problems.
In the months and years following the incident, Bangladeshi intelligence and police agencies cracked down on sleeper cells and extremist networks. Their primary focus was on capturing or killing high-profile targets associated with the Islamic State or a local Islamist terrorist network, Jamaat-ul-Mujahideen Bangladesh. Dozens of suspects were rounded up and arrested; 20 were eventually charged. At least seven of them were sentenced to death in Nov. 2019. Another eight suspects were killed during counterterrorist operations in different parts of the country.
To some observers, these heavy-handed operations appeared to be ruthlessly effective. Unlike most of its neighbors, Bangladesh has avoided large-scale extremist violence over the past four years. Yet beneath the surface, a different picture emerges. Online radicalization is widespread and increasingly normalized. A network of violent Islamist extremist groups are spreading hate speech and misinformation across the social media ecosystem. Bangladeshi police say that more than 80 percent of the people arrested for terrorism in recent years were radicalized online and almost 60 percent were university-educated.
There are fears that online rumors and misinformation could reignite a wave of extremism across Bangladesh. In fact, there are worrying signs that it is already starting. In early November, houses belonging to Bangladeshi Hindus were vandalized and torched after one of their occupants was accused of supporting France and defaming Islam on social media. The previous month, a librarian was lynched and killed after being accused of desecrating the Quran. The Bangladesh Peace Observatory reported a disturbing increase in attacks, police raids, and arrests in the first nine months of 2020 compared with the same period in 2019.
Worries about digital radicalization are hardly new. Some of the young men involved in the Holey Artisan Bakery attack were lively social media users and actively spread extremist content before the raid. They were also connected to other Bangladeshis living abroad, including one based in Canada, who helped plan and orchestrate the attack. Another convicted felon involved in helping plan the strike was first radicalized after watching videos produced by Anwar al-Awalki, a well-known Al Qaeda leader, as well as Jasimuddin Rhmani, a Bangladeshi preacher affiliated with al Qaeda who later aligned himself with the Islamic State.
But extremist content is growing in Bangladeshi cyberspace, much of it carefully packaged to evade detection by government and social media platforms. Examples include Jasimuddin Rahmani’s sermons on YouTube, which continue accumulating views. Lectures by Anwar al-Awlaki are routinely translated into Bangla and spread on Telegram, as well as recommended by sympathizers on Facebook. While sites sponsored by notorious groups such as the Islamic State are routinely taken down by social media companies, extremist propaganda in local dialects is available on specialized websites, discussion forums, social media channels, and even fact-checking sites where the curators add their own radical twist.
The spread of violent extremist ideology in cyberspace has real-world impacts. In recent years, there have been several incidents involving the targeted killing of liberal writers, digital influencers, and LGBTQ activists. Extremist groups affiliated with al Qaeda in the Indian Subcontinent (AQIS) are increasingly regarded as champions of socially conservative Islam. AQIS has an active presence on social media, though its followers generally avoid perpetrating staged acts of violence and tangling with Bangladeshi security agencies. Instead, they are stealthily targeting their enemies and normalizing hatred against “atheists” and “blasphemers.”
AQIS is disseminating divisive and intolerant narratives in Bangladeshi cyberspace. One of their goals is to persuade Muslims that “real” Islam does not support peace and tolerance. Instead, they are urging what they call an “intellectual jihad” in order to overcome the corrosive influence of infidels. They are waging a war of ideas, with hardline supporters exhorting followers to reject sanctions against polygamy, child marriage, and slavery. They are staunchly opposed to equal rights for women and routinely urge their followers to target blasphemers and secular liberals. AQIS is also working to discredit the government and international organizations, including their attempts to promote religious tolerance and moderate forms of Islam.
Owing to a combination of skillfully produced videos and viral content, AQIS is making impressive inroads among younger impressionable audiences. The sharing of violent extremist content—endorsements of armed jihad, glorification of terrorist groups, calls for attacks against liberal activists, and propaganda condemning minorities—is soaring. According to SecDev research, the number of interactions per Facebook post on violent extremist channels has surged by 250 percent in only one year, from an average of 99 interactions per post between April and June 2019 to 347 per post during the same months in 2020. This is partly due to the simultaneous growth of Islamist bloggers expressing support for AQIS.
Paradoxically, the COVID-19 pandemic has helped opportunistic extremist groups reach an even wider audience. Internet use exploded by over 50 percent across Bangladesh since the outbreak was first reported in early 2020. Violent extremists took advantage of increased connectivity to pull audiences into their webs of influence. In May 2020, just seven AQIS channels on YouTube featuring content related to COVID-19 registered a surge of over 100,000 new subscriptions. With the government distracted by the crisis, online radicalization has soared.
Domestic and international authorities are not taking the weaponization of social media lightly. Shortly after the 2016 Holey Artisan Bakery attacks, the Bangladeshi authorities joined forces with the United Nations to tackle violent extremism, both online and off. For example, they are working with a variety of entities such as the Bangladesh Peace Observatory to study the scope and scale of the challenge. With support from international partners, they also launched targeted programs to engage youth, including a digital peace movement involving over 1.9 million people.
Their aim is to inspire younger Bangladeshis to express solidarity online, while also bolstering the resilience of more vulnerable groups, especially minorities. The promotion of tolerance, inclusion, and digital literacy is the key to changing attitudes and, ultimately, behavior. Helping people of all ages to identify, question, and interpret what they read and share on social media is the first step in neutralizing divisive, exclusionary and ultimately hateful rhetoric designed to divide and polarize communities.
There are hopeful signs that these kinds of efforts will help diminish extremist content in Bangladeshi cyberspace. In recent months, the extent of disinformation related to COVID-19 has declined steadily. Dozens of extremist channels have been removed, although AQIS’s YouTube account still has over 600,000 subscribers. SecDev documented a 43 percent reduction of extremist content on roughly 400 social media channels between April and September 2020. Indeed, over 90 percent of Bangla channels affiliated with the Islamic State were removed between April and June 2020. That said, these kinds of channels are resilient and often reappear under new names soon after being shut down.
Mindful of the reputational risks, technology companies such as Facebook are stepping up their monitoring and strengthening content moderation policies. Facebook recently said that it now has fact-checkers monitoring posts in Bangla and has added “false information” tags to fake news and misinformation. Around 70 percent of the COVID-19-related disinformation posts identified by SecDev on Facebook have been removed or made inaccessible to the public. Other technology firms also appear to be moving to bolster their ability to detect and neutralize extremist material.
These challenges are likely to worsen before they improve. Social media use is expanding rapidly in Bangladesh and around the world. Efforts to solve the challenges of radicalization, misinformation, and disinformation must necessarily move beyond the whack-a-mole approach that still defines content moderation and account takedowns. Education, awareness-raising, and targeted investment in digital natives is ultimately the only way to fight violent extremism over the long term. This must be accompanied by a sustained debate about how to balance legitimate freedom of expression with the need for building a safer cyberspace.
