An influencer used artificial intelligence (AI) to create a deepfake of a politician, and fabricated a series of non-existent government policies – resulting in police involvement.
In early July, Malaysian content creator Tan Jian Hui uploaded an AI-generated clip on YouTube and TikTok depicting Chow Kon Yeow, chief minister of the Malaysian state of Penang.
Using footage from past interviews, Tan made a video of the chief minister talking about the Penang state government’s policy to “enable single men to find women.”
“We will allocate funds to primary schools and set up tutoring classes for kids… to nurture your future girlfriend,” said the AI-generated chief minister Chow.
Other ‘government policies’ mentioned by ‘AI-generated Chow’ in the clip include replacing Penang’s old street lights to LED-powered lights, setting up a new dialysis centre for patients, and launching a scholarship fund for students – all of which are false.
Bad timing
Less than 48 hours after Tan’s AI-generated video went viral on social media, Chow’s team filed a police report. The politician’s team claimed that the influencer’s move was a violation of rules set out by the Malaysian Communications and Multimedia Commission (MCMC).
“The influencer did not obtain my consent and used my past interviews for the video, which was immoral behaviour,” Chow told local Malaysian paper Oriental News Daily.
Chow also said that the influencer’s AI-generated clip constituted a criminal offence. “We hope that proper action can be taken after the police’s investigation,” Chow’s secretary, Teh Li Heng, told local media.
Tan’s deepfake video coincides with preparations for Malaysia’s upcoming state elections, where Prime Minister Anwar Ibrahim will face his first public vote of confidence amid an unstable government and a slowing economy.
AI-generated deepfakes
Video content made with AI is not new. However, with the arrival of tools like Midjourney, ChatGPT, and voice cloner ElevenLabs, creating has never been more accessible and time-saving, and laws are struggling to catch up.
In Australia, laws targeting deepfake content have yet to be drawn up, but victims could seek protection in existing defamation laws.
“If defamatory deepfakes are created and published, the victim of the deepfake may have a claim against anyone involved in the publication of the deepfake to compensate for the damage to the victim’s reputation,” stated Sydney law firm White Knight Lawyers.
After much public outrage, Tan finally uploaded a video on YouTube apologising for the AI-generated deepfake clip.
“I just wanted to make a video to share with everyone how Penang has changed…I am disappointed in my oversight and I regret any pain or stress I caused to the chief minister. I assure you that something like this will never happen again,” Tan said.
