Attention Economy and Rising Social Polarisation

A few years ago, social media felt largely like an extension of our social lives. We found old friends, shared photographs, discussed ideas and discovered information that traditional media might have missed. Today, opening Facebook, TikTok, YouTube or X can feel very different. Political arguments, anger, accusations and emotionally charged videos compete continuously for our attention. The technology has changed, but so has the economic model behind it. Now, most social-media platforms make money by keeping us engaged. The longer we watch, click, react, comment and share, the more commercially valuable our attention becomes. This is commonly called the “attention economy”.

Nevertheless, that does not mean social media inevitably drives extremism. The evidence is more nuanced, but concerns remain about platform incentives. A recent economic model found that giving more algorithmic weight to likes and shares boosts engagement while also increasing misinformation and polarisation.

A large cross-country study published in Nature Communications examined political interactions on Twitter (now X) across nine countries, including the United States, Britain, France, Germany, Spain and Turkey. It found polarised political networks and greater toxicity towards political out-groups. This matters because anger is no longer just an emotion expressed online. In the attention economy, it can also be economically valuable. The consequences are increasingly visible across countries, and countries are responding differently. 

Australia has adopted one of the strongest child-protection approaches. Since December 2025, age-restricted social media platforms have been required to take reasonable steps to prevent Australians under 16 from maintaining accounts. The rules apply to Facebook, Instagram, TikTok, YouTube, Snapchat and X, and target features such as algorithmic recommendations, infinite scrolling, visible likes and disappearing stories that can drive prolonged engagement. Non-compliant platforms can face penalties of up to AUD109.2 million. The European Union has taken a different approach through the Digital Services Act. Rather than focusing primarily on users, it places obligations on large technology companies to address systemic risks, increase transparency and oversee recommender systems. European regulators have, for instance, examined whether the recommendation systems of Facebook and Instagram meet these requirements.

These approaches raise an important question: should governments concentrate only on what people post, or should they also examine the systems deciding what millions of people are encouraged to see? For Bangladesh, that question is increasingly relevant. Bangladesh had an estimated 82.8 million internet users and 64 million social-media user identities at the end of 2025. Social-media identities were equivalent to around 36 per cent of the country’s population. The figures do not necessarily represent 64 million unique individuals, but they show the enormous scale of Bangladesh’s digital public space. For millions of Bangladeshis, Facebook and other platforms have effectively become a combination of newspaper, marketplace, political discussion room and neighbourhood meeting place.

There are enormous benefits. Small businesses can reach customers cheaply. Citizens can discuss public affairs. During emergencies, information and requests for assistance can travel quickly. People who once had little access to traditional media can participate in national conversations. But the same infrastructure can distribute rumours, hatred and fabricated information at extraordinary speed.

Bangladesh has already faced this challenge in politics. Research by the University of Liberal Arts Bangladesh examined disinformation and misinformation on social media during the January 2024 national election. But political misinformation is only one concern. False allegations involving religion, ethnicity or individuals can be especially dangerous in a densely populated society, where online outrage can quickly trigger offline conflict.

Greater government control over speech is not the answer. Bangladesh’s Cyber Security Act 2026, which replaced the 2025 Cyber Security Ordinance, provides a legal framework for cyber offences. Yet regulating speech alone does not address the deeper attention-economy problem: the platform systems that determine what gains visibility. Protecting social cohesion should not come at the expense of legitimate criticism, journalism or political dissent. Instead, Bangladesh should focus not only on speakers, but also on holding platforms accountable for the systems that amplify speech.

Major platforms could be required to produce Bangladesh-specific transparency reports showing how much Bangla-language harmful or illegal content is reported and removed, how quickly credible complaints are addressed, and how users can appeal decisions. Researchers and universities should have privacy-protected access to relevant platform data so that misinformation and polarisation can be independently studied.

Algorithmic transparency and accountability deserve particular attention. Platforms should explain in understandable terms why content is being recommended. Users could be given a genuine choice between algorithmically selected and chronological feeds. Platforms could also introduce more friction when potentially misleading material suddenly becomes viral, for example by asking users to open an article before sharing it.

Children need stronger protection too. Bangladesh does not necessarily need to copy Australia’s under-16 model, but Australia’s experiment deserves close observation. Restrictions on targeted advertising to children, stronger privacy defaults, parental tools and safeguards against addictive design features are reasonable areas to examine.

Regulation alone will not solve the problem. Digital and media literacy should be part of mainstream education. Students should learn to verify sources, identify manipulated images and AI-generated content, distinguish reporting from opinion and understand how algorithms shape what they see online. Similar training can be provided to teachers, journalists, public officials, religious leaders and community organisations.

Bangladesh could also establish an independent multi-stakeholder mechanism involving technology companies, universities, journalists, fact-checkers, civil society and regulators. It could identify emerging digital risks and recommend responses while safeguarding freedom of expression.

Ultimately, the challenge goes beyond misinformation. A healthy society requires people to disagree without treating one another as enemies. Social media did not create Bangladesh’s political, religious or social divisions, but technology can greatly increase their speed, visibility and intensity. The goal is not a quieter internet through censorship, but a healthier digital public space where technology does not systematically favour outrage over reason simply because outrage attracts clicks. Attention has become an economic commodity. Bangladesh must ensure its social cohesion does not become the price paid for it.

________________________________________

The writer is an academician and development strategist, currently serving at a foreign diplomatic mission in Dhaka. He can be reached at [email protected]