In Sri Lanka’s fiercely contested digital landscape, disinformation and misinformation have emerged as formidable threats to democracy, social harmony, and public trust. 

Recent instances of misinformation, and possibly disinformation, are reports that Prime Minister Harini Amarasuriya had visited former President Ranil Wickremesinghe in hospital and a claim that British High Commissioner Andrew Patrick attended Wickremesinghe’s court hearing, both of whom have denied the allegations publicly. 

Darshatha Gamage, Director of Impact Voices, a research and advocacy organisation specialising in civic technology and climate change, underscored the potential of dedicated Artificial Intelligence (AI) tools to support content verification, especially within newsrooms and fact-checking entities. 

“Social media platforms can leverage AI to detect patterns and behaviours of actors who consistently disseminate disinformation,” he noted. However, the challenge surfaces in fast-moving information environments — such as breaking or live news — where AI systems depend on the availability of verified data to be effective. “AI can only curb the spread of misinformation if credible sources make reliable information accessible in real time,” he added.

Gamage also highlighted AI’s anticipatory capabilities, revealing that it could analyse emerging social media narratives to intervene before disinformation escalated. “Where images are involved, AI enhances verification by detecting anomalies, examining metadata, and identifying watermarks,” he noted. Thereby, it enables platforms to label manipulated or AI-generated content more transparently. 

Still, he cautioned: “AI interventions should not undermine freedom of expression on social media, except where a clear threshold of harm is met.”

AI’s mixed role

Despite AI’s promise, a significant limitation exists in Sri Lanka: the lack of locally relevant data processing and model training, which restricts the accuracy and efficacy of AI systems designed to tackle disinformation. 

This gap underscores the urgent need for tools that understand local languages and contexts, a gap currently being addressed by initiatives such as LIRNEasia’s ‘Dissect’ tool. 

This AI-powered fact-checking web tool, compatible with Sinhala and in development for Tamil, assists journalists and fact-checkers by breaking down content into claims and sources without attempting to classify truth or falsehood, thus providing a critical aid in navigating the complex information landscape.

Media analyst Nalaka Gunawardene reiterated these nuances about AI’s mixed role. 

While recognising the many positive applications of Generative AI (GenAI) in education and research, he warned of its dark side: “GenAI tools make it much easier to fabricate images, videos, and audio of events that never happened.” 

Deepfakes — the highly realistic synthetic media designed to impersonate individuals — are a notable example, weaponised by disinformation originators to spread falsehoods convincingly.

Gunawardene emphasised that AI could also support content moderation and fact-checking at a systemic level, aiding platforms to spot patterns and false information. 

However, he stressed serious limitations, particularly the inadequate development of AI tools for content moderation in Sinhala and Tamil, Sri Lanka’s main social media languages. “Global platforms should not be unilaterally deciding what is true or false,” he cautioned, highlighting the complex question: who are the arbiters of truth in today’s multi-narrative societies? 

There are “no easy answers or quick fixes,” he concluded, reflecting the profound challenges in digital spaces marked by competing perspectives.

Flaws of the Online Safety Act 

Both Gamage and Gunawardene fiercely critiqued Sri Lanka’s Online Safety Act (OSA) No.9 of 2024, which aims to combat disinformation but falls short in key respects. 

Gamage noted: “On paper, the OSA could address disinformation by recognising it as the dissemination of false statements. However, the act fails to clearly define what constitutes a false statement, leaving significant room for subjective interpretation.” He warned that the law’s broad powers granted to the politically appointed commission raised serious concerns about impartiality.

“The law operates largely as a reactive mechanism and by the time the commission reviews complaints, false narratives have often already spread widely, reducing intervention effectiveness.” Gamage expressed concerns that the act risked being weaponised against free speech and dissent rather than genuinely safeguarding against disinformation.

Gunawardene’s commentary is even more critical. Calling the act “deeply flawed,” he labelled it a case of “the remedy being far worse than the ailment”. He traced its origins to a rushed parliamentary process lacking accountability and public consultation and noted that even those who had initially opposed the bill now contemplated amending it instead of outright repeal. 

“The OSA is flawed for many reasons, mainly due to vague definitions and overbroad provisions,” he argued. “Falsehood-based offences easily become tools for prosecutors to silence inconvenient questions or unpopular opinions within constitutional free speech boundaries.”

Gunawardene also expressed concern regarding the act’s lack of precise truth definitions, noting that it risked privileging ruling-party narratives or majority views while punishing dissent. He cited the example of the politically promoted but scientifically dubious ‘Dhammika Paniya’ herbal tonic during the Covid-19 pandemic, demonstrating the complexities of what constitutes ‘truth.’ This ambiguity exposes disinformation laws to misuse, stifling critical journalism and public discourse.

He stressed that criminalising disinformation with vague laws conflicted with human rights commitments and called for caution to avoid direct or indirect censorship, suppression of critical thinking, or rights violations.

Building a culture of responsibility  

Beyond technological and legal remedies, Gamage advocated fostering a culture of responsibility around information sharing as a crucial civic duty. He stressed developing digital literacy at all societal levels — schools, universities, workplaces, and communities — training individuals to question sources, detect manipulative narratives, and appreciate the harm of spreading unverified claims.

“When people understand how misinformation impacts democracy, public safety, and social trust, they are more likely to pause before clicking ‘share,’” he said. Community voices, he insisted, were powerful forces encouraging fact-checking as a civic habit. Gamage highlighted that consumers of media must demand verified information, holding not only media organisations but also content creators and ordinary online users accountable.

He underscored the pivotal role of independent fact-checking organisations, which must follow rigorous standards such as those set by the International Fact-Checking Network (IFCN) to ensure credibility. Crucially, he stressed, fact-checking must remain independent of the State to avoid political influence. Past experiences with Government-run fact-checking efforts in Sri Lanka have proven ineffective and politically driven.

However, the State should support independent fact-checkers financially and through facilitating collaboration with social media platforms, holding these platforms accountable for curbing misinformation.

Gunawardene similarly called for multiple responses to disinformation beyond digital literacy, such as enhanced Government transparency, proactive disclosure, compliance with the Right to Information Act, and robust media ethics in both State-owned and private media. He criticised the heavy use of State media for propaganda, calling for editorial independence to increase the supply of credible information. 

He further emphasised the need for civil society scaling up independent, non-partisan fact-checking and pressing global tech platforms like Meta, Google, and TikTok to improve content moderation in Sri Lanka’s local languages.

Both experts agreed that many Sri Lankans lacked critical thinking skills needed to evaluate information credibility. Gunawardene warned: “Most of us don’t know how to critically consume information that reaches us. Most who use smartphones are not half as smart as their phones.” 

This digital literacy gap is a gateway for disinformation — a problem compounded by eroding trust in mainstream media, which citizens often see as biased or unreliable. This credibility vacuum drives users towards alternative sources with weaker editorial standards, creating fertile ground for false narratives.

Need for strengthened fact-checking ecosystem

The scarcity of credible, visible fact-checking organisations in Sri Lanka is a shared concern. Gamage stressed the need to both establish new bodies and reinforce existing ones, endowed with adequate resources and independence. An expanded fact-checking ecosystem equipped with rigorous standards is vital to empower citizens and counter misinformation effectively.

Gunawardene echoed that fact-checking must not be a State function, given that State agencies themselves sometimes generated disinformation. He warned against public funds being wasted on Government-run fact-checking that lacked credibility or efficacy. Instead, public resources should support independent efforts. 

He also highlighted the potential impact of improved editorial independence at State-owned media outlets, which had long served as ruling party propaganda tools.

Collective effort against disinformation

Sri Lanka’s struggle against disinformation is multifaceted, involving advanced AI technologies, flawed legislation, media ethics, civic responsibility, and education reform. 

“Spreading false information should be recognised as an act that undermines collective safety and social harmony,” Gamage asserted, noting that it should be made a shared duty among citizens, not just governments and platforms.

Gunawardene summed up the complexity: “There are no short-term solutions. In an open democratic society, no single actor should control what is true or false. Multiple actors, along with government transparency, media responsibility, civil society fact-checking, technology platforms, and empowered citizens, must work in concert.”

When The Sunday Morning reached out to the Media Ministry to determine what mechanisms were in place to address the issue, it was revealed that there was no programme or authority in place exclusively for the purpose of tackling misinformation and disinformation. 

Ministry of Media Additional Secretary Kelum Wijenayake said that the ministry itself did not have the apparatus and that the responsibility for cybersecurity was split between the Sri Lanka Computer Emergency Readiness Team (SLCERT) and the Police. “SLCERT handles matters pertaining to direct cyberthreats like hacking while the Criminal Investigation Department (CID) deals with curtailing the spread of false information online.”

SLCERT Chief Information Security Officer Nirosh Ananda said that a private organisation played a part in tackling the propagation of false information online. “SLCERT deals solely with cybercrimes pertaining to Government websites. We do not deal with mis/disinformation. To my knowledge, the CID as well as a private organisation known as FactCheck Sri Lanka are sharing that responsibility.”

As technology evolves and digital platforms multiply, Sri Lanka faces an urgent imperative to strengthen local AI tools, reform laws that risk censorship, enhance media integrity, and cultivate a digitally literate, critically thinking citizenry. This systemic approach is essential to safeguard the country’s democracy, foster social trust, and build resilience against the growing tide of disinformation.