Safeguarding Truth and Cultivating Critical Minds in a Machine-Generated World

Access to information was considered the cornerstone of human development for several hundred years.

The Enlightenment argued that knowledge would be broader, making citizens more enlightened and committed to democracy, strengthening democratic institutions, and getting social development moving faster.

But today, the problem is not skimp on information as overwhelming abundance of it.

Artificial Intelligence (AI) has turned the once purely digital world into an instant world of text, images, audio, and videos with tremendous realism. So, the key question of our time is no longer accessing information but rather separating fact from fiction.

AI has quickly woven its way into the fabric of daily life, impacting how human and artificial intelligence have influenced education, after-school programs, healthcare systems, financial markets, governance, and journalism.

Large language models and generative AI systems can give you a summary of the report, translate multiple languages in seconds, analyse data instantly, and write persuasive content.

Though they have enhanced productivity and expanded access to knowledge, these innovations come with unprecedented risks. AI has the ability to create real-looking photos, fake human voices, realistic news stories and videos that look real, yet are totally made up.

With the decomposition of content, the epistemic crisis faced by citizens continues to prevail: what can be believed, who is manipulative, and who is entirely fake as a function.

Scholars refer to this phenomenon as epistemic fragility, the state that occurs when technological systems damage society’s capacity to generate, assess, and produce trustworthy knowledge.

When misinformation can be manufactured in bulk and spread around the world in minutes, public trust in journalism, scientific research, democratic institutions, and even video footage starts to trickle away.

So, the difficulty is not simply technological but inherently epistemological, as it pertains to the processes by which societies establish and maintain truth.

Such a setting must acknowledge media literacy as more than merely a digital skill. This is a crucial type of epistemic defence. Media literacy teaches people the skills to appraise sources, spot bias, identify manipulation and ascertain whether evidence is verifiable, amongst other things, so they are equipped to question information instead of accepting or sharing.

Instead of passively consuming algorithmically filtered content, media-literate citizens function as active evaluators who can fend off malicious information and make informed choices.

This need has been reiterated by international organisations on several occasions. In its Global Risks Report 2024, the World Economic Forum highlights misinformation and disinformation as one of the highest-impact global risks in the short term, arguing that they pose a key threat to democratic stability, public trust and social cohesion.

In the same way, UNESCO’s AI and Education: Guidance for Policy Makers (2023) states that we need to make critical digital literacy a primary item on the educational agenda so that citizens can engage in media environments increasingly mediated by AI responsibly.
Bangladesh is at a critical juncture of this transition.

National Artificial Intelligence Policy 2026–2030 (Draft V2.0)reflects an ambition of making use of AI to ramp up economic development and digital governance, innovation, and also through public service delivery. However, the policy discussion has revealed significant issues around transparency, accountability, ethics, and public trust.

In Bangladesh, there have been amazing strides in connectivity and digital inclusion, but there is still a long way to go on media literacy. In rural communities, though, the opportunities for digital learning are scarcer overall; urban areas typically have better access to both educational resources and digital infrastructure.

Further, curricula tend to emphasize basic technological skills but place significantly less emphasis on AI-generated content, algorithmic bias, source verification, or critical reasoning. Despite having high levels of digital connectivity, many users lack the skills necessary to traverse increasingly complicated information ecosystems.

Human psychology increases this vulnerability drastically. People are prone to compelling but deceptive narratives because they naturally depend on cognitive shortcuts and emotional reactions while processing information.

Misinformation in Bangladesh exacerbates political polarization, religious sentiment, and social discord. Because emotive content increases attention on digital platforms, sensational accusations frequently spread faster than meticulously confirmed reporting.

These threats are made worse by the advent of deep fakes and synthetic media. A Nature Human Behaviour study found that even well-educated individuals were unable to easily differentiate AI-generated images from real photographs, suggesting that technological advancement can quickly overtake human vision.

Recommendation algorithms are an additional problem that influences them significantly these days. Social media features also prioritise what users will see, leading to what value their content may have as opposed to whether it is actually true. Content that provokes outrage, fear, or strong emotional reactions is often amplified because it generates higher levels of interaction.

Gradually, through these systems of algorithmic reinforcement, people are placed in echo chambers while reducing exposure to diverse perspectives. So much so that for millions of Bangladeshis who use Facebook, YouTube, and TikTok as the main sources of news nowadays, algorithmic curation has become a hidden influential pillar of public opinion.

However, in the age of generative AI, media literacy is no longer sufficient; AI literacy is now necessary. In addition to critically analysing media, citizens need to comprehend how recommendation algorithms trap users in information echo chambers, how AI systems process data to produce text, and the inherent limitations of huge language models that necessitate human verification.

AI should enhance human judgment rather than replace it. From journalism to education, business and public administration, human oversight must be at the core of decision-making.

Solving such challenges will require a concerted push at the national level. Media and AI literacy should be included in school, college and university curricula as fundamental competences rather than elective courses.

Teacher training programmes should train educators in assessing digital efficacy to teach critical assessment of AI and its usage.

Strengthening editorial transparency, scaling up independent fact-checking initiatives, and improving communication about verification practices that media organisations undertake.

On the technology front, tech firms should take greater responsibility in lowering demonstrably false information from being amplified algorithmically while preserving legitimate freedom of expression.

Policymakers should develop AI governance frameworks that are technologically focused; there needs to be equal emphasis on educating the public.

Media literacy also has an important economic impact. Employers are also actively searching for these professionals who can interpret information critically, evaluate evidence, and make decisions in digitally mediated workplaces as AI reshapes labour markets.

Enhancing these competencies will not only bolster democratic resilience, but it will also help ensure the competitiveness of the workforce and sustained economic productivity in an expanding digital economy for Bangladesh.

Ultimately, combating misinformation is about preserving society’s ability to understand. Democratic societies maybe unsustainable if people lose faith in evidence, experts, and objective reality. Technology such as artificial intelligence will continually transform how information is consumed, distributed, and created, but the future of truth cannot be decided by technology. This depends on people who have knowledge and can think critically.

(Rashadul Islam: A Freelance Contributor to National & International Dailies, Columnist and Independent Researcher, Affiliation: University of Barishal and Ishtiaq Mainuddin, Lecturer, Department of Accounting & Information, stems, Bangladesh Maritime University.)