Sri Lanka's President Never Endorsed That Investment Plan. A Deepfake Did.
What happened
A video circulating in Sri Lanka in June 2025 showed the country's president, Anura Kumara Dissanayake, endorsing what appeared to be a government-backed investment opportunity. He had not. The footage was fabricated, a deepfake built from real television material and designed to extract money from people who had no way to tell the difference.
The operation was constructed from at least two distinct elements. Scammers took footage from a May 3rd broadcast of the Satana TV program, apparently a legitimate appearance featuring the president, and layered over it a cloned or dubbed English-language voice endorsing the scheme. Alongside the video, they distributed a fake news article designed to look like content from the Daily Mirror, a recognized Sri Lankan English-language publication. The combination was deliberate: a recognizable face, an institutional-sounding voice, and a credible press outlet, all pointing toward the same fraudulent sign-up link.
The fraud was identified relatively quickly. Sri Lanka CERT, the country's computer emergency response team, flagged the video as AI-generated. Journalists who investigated confirmed the footage was manipulated. The Presidential Media Division also issued a denial, publicly confirming the video did not depict anything the president had actually said or authorized. That three separate institutions had to issue public corrections illustrates how much labor is required to respond to a single, targeted piece of synthetic media.
What the incident shows about the current threat landscape is not surprising but is worth naming directly. A head of state's likeness and voice can be cloned well enough to serve as the anchor for a financial fraud operation. The materials required are a few minutes of real broadcast footage and access to voice synthesis technology. The fake news article adds a second layer of apparent credibility, and together the package targets people who would naturally trust an investment plan their government appears to be promoting.
The accountability gap here is not a technical one. Detection tools exist in research settings and are improving. What is missing is any requirement that content platforms, distribution networks, or the people releasing synthetic media leave a provable record of what a system produced and when, something that could be checked against the original footage before the content reaches its targets. Without that record, the burden of disproving a deepfake falls entirely on the institutions being impersonated, after the fraud is already in motion.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.