Artificial intelligence is rapidly reshaping the information landscape across South Asia, and Sri Lanka is increasingly finding itself at the center of a troubling new frontier. A newly released report has sounded urgent alarms about the growing proliferation of AI-generated synthetic personas and political deepfakes circulating on Facebook, warning that these digital deceptions pose a serious and escalating threat to democratic processes, public trust, and social cohesion in the island nation.
What the Report Reveals
The report, which focuses specifically on AI-generated content spreading across Facebook in Sri Lanka, highlights two particularly alarming trends. First, the emergence of synthetic female personas — artificially created profiles featuring AI-generated images of women — being used to build audiences, spread narratives, and lend false credibility to misleading content. Second, the increasing use of political deepfakes, which involve manipulated audio and video content designed to make real politicians appear to say or do things they never actually did.
Researchers found that these synthetic accounts are becoming increasingly sophisticated, making them difficult for ordinary users to identify as fake. The AI-generated profile images used are often indistinguishable from real photographs, and the accounts frequently engage in coordinated behavior to amplify specific political messages or discredit public figures. This level of deception represents a significant evolution from the crude misinformation tactics observed in previous years.
Why Sri Lanka Is Particularly Vulnerable
Sri Lanka presents a uniquely challenging environment when it comes to combating AI-driven misinformation. The country has experienced significant political instability in recent years, including the dramatic 2022 economic crisis that led to widespread public protests and the ousting of then-President Gotabaya Rajapaksa. This volatile political climate creates fertile ground for bad actors seeking to exploit public anxieties and divisions.
Additionally, Sri Lanka's multilingual population — spanning Sinhala, Tamil, and English speakers — creates additional complexity for content moderation efforts. Automated detection systems often struggle with regional languages and dialects, allowing harmful AI-generated content to slip through platform safeguards more easily than it might in predominantly English-speaking markets. With smartphone penetration continuing to grow and Facebook remaining the dominant social media platform for news consumption, the potential reach of deepfake content is enormous.
The Gendered Dimension of AI Manipulation
One of the most striking findings of the report is the deliberate use of AI-generated female personas as vehicles for political manipulation. This tactic exploits deeply embedded social dynamics, as profiles featuring women may be perceived as more trustworthy, emotionally relatable, or less overtly political than those featuring male personas. By constructing elaborate fake identities around AI-generated women, those behind influence operations can build substantial followings before deploying these accounts to push targeted political messaging.
This gendered approach to synthetic persona creation also raises serious concerns about the broader objectification and weaponization of women's likenesses in digital spaces, even when the women depicted do not actually exist. Experts warn that normalizing the use of fabricated female identities in this manner contributes to a wider culture of digital manipulation that ultimately harms real women, particularly those in public and political life who may find their own images targeted by deepfake technology.
Political Deepfakes and the Threat to Democracy
Beyond synthetic personas, the report draws particular attention to the danger posed by political deepfakes targeting Sri Lankan leaders and public figures. Fabricated videos or audio clips depicting politicians making inflammatory statements, announcing false policy positions, or engaging in scandalous behavior have the potential to rapidly destabilize public discourse, especially when shared through trusted social networks before fact-checkers can respond.
In a country where political tensions between ethnic and religious communities remain sensitive, a well-crafted deepfake targeting a prominent leader could trigger real-world consequences with alarming speed. The report urges both platform providers and government authorities to treat this threat with the seriousness it deserves, calling for stronger detection mechanisms, faster response protocols, and greater public awareness campaigns.
Calls for Action and What Comes Next
The report concludes with a series of recommendations directed at multiple stakeholders. Social media platforms, particularly Facebook's parent company Meta, are urged to invest more heavily in AI detection tools capable of identifying synthetic content in regional languages and to improve transparency around content removal decisions. Civil society organizations are called upon to expand digital literacy programs that help ordinary Sri Lankans recognize and critically evaluate potentially manipulated content.
Government bodies are encouraged to develop clear legal frameworks addressing the creation and distribution of harmful deepfakes without infringing on legitimate free expression. Regional cooperation with neighboring countries facing similar challenges is also highlighted as a critical component of any effective long-term strategy.
As artificial intelligence tools become cheaper, more accessible, and increasingly powerful, the window for establishing meaningful safeguards is narrowing. Sri Lanka's experience serves as a cautionary tale and a call to action for democracies across the developing world grappling with the same converging threats of technology, political polarization, and information disorder.