SINGAPORE – To spot a deepfake video, look out for distorted images, speech that does not match the movement of the speaker’s lips, and claims that are too good to be true.

Tech experts offered these tips as they called for more education to detect fake videos made using artificial intelligence (AI) after several deepfakes surfaced in December that involved public figures like Prime Minister Lee Hsien Loong.

PM Lee on Dec 29 warned the public in a Facebook post not to respond to scam videos on investments or giveaways after a deepfake video of him purportedly promoting an investment surfaced. It is not known who was involved in the misinformation campaign.

Deputy Prime Minister Lawrence Wong, whose appearance has also been used to promote investment scams, alerted the public on Dec 11 of deepfake posts spreading misinformation that the authorities were planning a circuit breaker amid a spike in Covid-19 cases.

The likeness of Madam Ho Ching – Mr Lee’s wife and former Temasek chief executive – was also manipulated by fraudsters in another investment video that surfaced in December.

The video circulated as an advertisement on YouTube, which plays before users can view the videos they have selected. It is not known who uploaded the false ad.

A spokesman for Google, which runs YouTube, said it has taken down the flagged ad and terminated the associated channel.

Google did not state whether the ad was vetted before it was approved, but said that YouTube will soon require content creators to disclose when they have created realistically altered or synthetic content using AI or other tools. This will be made clear to viewers in the video’s description panel, and labelled on the video itself for sensitive content, such as those related to elections, conflicts or public health crises.

Source link

By admin

Leave a Reply

Your email address will not be published. Required fields are marked *