What this lesson teaches
I can explain how individuals can stay alert and help keep their country safe.
Syllabus K336, Issue 3, Guiding Question 4, Responses to security impacts of globalisation. Vigilance by individuals
Make a guess
In December 2023, videos showed Singapore's leaders promoting an investment scheme. Do you think a video of a real person speaking must be real?
- Yes, if it has been shared by many people you trust online.
- No, AI deepfakes can fake what a real person says.
- Yes, because you can see and hear the real person.
Show the answer
No, AI deepfakes can fake what a real person says.
The leaders warned the public that the videos were fake scams. Checking before sharing is the first defence.
Fake content spreads across borders in seconds. Laws help, but each person checking before sharing is the first defence.
Misinformation is false information. When it is spread on purpose to mislead, it is called disinformation. A deepfake is a video, picture or recording made or changed with AI so that a real person appears to say or do something they never did.
Fake content threatens security because it can cause panic, cheat people of money or turn groups against each other. It can be made anywhere in the world and shared in Singapore within minutes.
In December 2023, deepfake videos of then Prime Minister Lee Hsien Loong and then Deputy Prime Minister Lawrence Wong were used to promote investment scams. PM Lee warned the public not to believe such videos.
Singapore's laws include the Protection from Online Falsehoods and Manipulation Act (POFMA), in force since 2 October 2019, which allows the government to order corrections to false statements. In October 2024, Parliament passed a law banning realistic deepfakes of election candidates during elections. But laws act after content spreads, so individuals still need to stop, check and not forward.
- Deepfake
- AI-made or AI-changed media showing a real person saying or doing something they never did.
- POFMA
- In force since 2 October 2019; allows orders to correct online falsehoods.
Real examples to use
Election deepfakes law
Passed on 15 October 2024; bans realistic deepfakes of candidates from the issue of the Writ of Election to the close of polling.
Worked example: Using this briefing in an answer
A Question 6 asks how individuals can help keep Singapore safe from transnational threats.
- Identify: individuals can stay vigilant by checking content before sharing it.
- Example: in December 2023, deepfake videos showed then PM Lee Hsien Loong promoting an investment scam.
- Outcome: if people check official sources first, fakes reach fewer people, so fewer are cheated or misled.
- Link: this protects trust in public information, which keeps society calm and united during a crisis.
Watch out for this
If a video shows a real person speaking, it must be real.
AI can copy a person's face and voice. Check where the video came from and whether official sources report the same thing.
Check your understanding
What makes a deepfake different from other false posts?
- It is made by foreign governments to sway how people vote.
- AI makes a real person seem to say or do it.
- It spreads faster than other false posts.
Show the answer
AI makes a real person seem to say or do it.
Right. A deepfake fakes a real person's face or voice, which makes it more convincing.