Singapore has proposed new legislation to outlaw the use of digitally manipulated content, specifically deepfakes, during elections.
The new legal steps will safeguard “the integrity and truthfulness of representation” during elections, said the Ministry of Digital Development and Information (MDDI) in a statement on Monday. The proposed bill was introduced in parliament and will be tabled for a second reading at the next sitting.
It aims to provide new legislation to safeguard the local population against digitally manipulated content, including audio, video, and image generated by artificial intelligence (AI), otherwise known as deepfakes.
Also: Businesses still ready to invest in Gen AI, with risk management a top priority
Advancements in generative AI (Gen AI) technology have opened up opportunities, but also present information risks, with threat actors leveraging it to create and spread misinformation, MDDI said. It pointed to an increase in malicious deepfakes used in scams and extortion in Singapore.
While online falsehoods can be dealt with through the country’s existing Protection from Online Falsehoods and Manipulation Act, “targeted levers” are necessary under its Online Election Advertising Act to address deepfakes that misrepresent candidates, the ministry said. The latter legislation defines any content accessible online that can be “reasonably” deemed to promote or enhance the standing of political parties or candidates or prejudice the electoral prospects of other political parties or candidates.
If left unaddressed, deepfakes can compromise the integrity of Singapore’s electoral process, MDDI said. It added that voters should be able to make informed decisions based on facts, not misinformation.
Also: Artificial intelligence, real anxiety: Why we can’t stop worrying and love AI
The proposed legislative measures will prohibit publishing digitally generated or manipulated online election advertising that “realistically depicts” candidates saying or doing something they did not say or do. This includes content that uses both AI and non-AI techniques, such as editing via photo editing, dubbing, or splicing.
The new measures will apply only to online election advertising that features individuals running as candidates in the election.
The proposed legislation will allow corrective directions to be issued to individuals who publish such content as well as social media platforms, to take down offending content. Directives can also be sent to internet service providers to disable access by Singapore users to such content during the election period.
<!–>
Failure to comply with corrective directions has been deemed an offense and is punishable by a fine, imprisonment, or both if convicted.
The bill also will allow candidates to submit requests for content to be assessed for potential breach and corrective directions issued accordingly. Candidates who have been misrepresented by such content can make a declaration to attest to the veracity of their claim, MDDI said.
It will be illegal for candidates to knowingly make false or misleading declarations in their request on the content in question, the ministry said. It added that a fine or other penalties may be meted out should candidates be found to have committed such practices.
Also: Can governments turn AI safety talk into action?
The Singapore government also plans to introduce a code of practice mandating specific social media providers to implement measures to prevent and combat the use of digitally manipulated content on their platforms.
Details for this will be established in the coming months, with discussions to take place between the regulator and social media operators, MDDI said.
“Beyond elections, deepfakes and other forms of manipulated content can pose serious harm to online users,” it said. “Creators and distributors of AI-generated content have a responsibility to foster a safe information space, both during and outside of elections.”
Also: How Singapore is creating more inclusive AI
Singapore is anticipated to hold its general elections early next year or before November 2025.
South Korea adopted a 90-day ban on the use of deepfakes in political campaign content for its April 2024 legislative election.
