Ahead of the South Korean general elections, there is a spike in deepfake instances. What are the problems?

Election officials are being extra cautious since there are fewer than 50 days left before South Korea’s parliamentary elections in April 2024. These worries stem from the widespread use of deepfake pictures and videos produced by artificial intelligence (AI).

The nation’s election watchdog, the National Election Commission (NEC), reports that from January 29 to last week’s end, 129 instances of artificial intelligence (AI)-generated media material were found to be in breach of the recently amended election legislation.

According to Yonhap news agency, this rule forbids the use of deepfakes in political campaigns and carries fines of up to $10 million won ($7,500) or seven years in jail for violators.

The NEC’s offensive against deepfakes is a proactive reaction to the way misinformation is changing, helped along by advances in artificial intelligence.

The National Assembly amended the law in December with the intention of preventing the dissemination of misleading information and protecting the voting process’s integrity from new dangers.

Previous elections have been manipulated using deepfake techniques, both nationally and globally. This has led to fears that artificial intelligence-generated films and pictures pose a serious danger to democracy, particularly when it comes to the spread of false information and swaying public opinion.

An AI-generated video, for example, appeared on social media during the 2022 provincial elections, showing President Yoon Suk Yeol supporting a local candidate from the governing party.

Voters in New Hampshire were advised not to cast ballots in the state’s presidential primary by an AI-generated robocall that mimicked US President Joe Biden and was sent in January.

Experts have cautioned about the increasing complexity and speed of deepfake manufacturing, which is surpassing the conventional verification methods used by electoral officials, as AI technology continues to advance. The unparalleled speed at which deepfakes are created in comparison to verification attempts makes it difficult for authorities to stop the spread of these fakes.

Professor Kim Myuhng-joo of Seoul Women’s University said, “The speed of production of believable fakes through deepfake technologies is much faster than that of the NEC’s verification process.” “It’s unmatchable.”

Moreover, the ongoing development of deepfake technology presents new challenges for detection and mitigation, necessitating a multipronged strategy that places a high priority on stringent verification methods.

Related Articles

Back to top button