(Subscribe to our Today's Cache newsletter for a quick snapshot of top 5 tech stories. Click here to subscribe for free.)
Synthetic media is growing exponentially with malicious actors targeting individuals most connected to the entertainment business, followed by politics.
As of June 2020, deepfake videos identified online doubled to 49,081 in just six months since January, according to Deep Trace Lab, a deepfake detection technology firm.
The firm’s analysis points to a rising number of synthetic content that mimicks people in the entertainment industry. Since July 2019, about 95% of all the deepfake videos identified by Deep Trace had their primary targets in the film, fashion, sports and media industries.
In October 2019, a YouTube channel made a synthetic video of a scene from the movie Matrix. The scene was altered to replace Keanu Reeves’ face with Will Smith’s when the character Neo is asked to choose the red or green pill.
Most recently, in May, the same channel edited a scene from the famous movie Indiana Jones, and face-swapped Harrison Ford with Chris Pratt.
Deepfake video editors aren’t limiting their skills to famous Hollywood actors. About two-thirds of their target are on Instagram, Twitch and YouTube.
Twitch streamer Mizkif’s face was reportedly deepfaked onto singer Sam Smith’s by a fan of the streamer.
West to East
On the eastern side of the globe, China saw its first appearance of Deepfake in Chinese web TV series this year.
Following a clash with the local authorities in the country, actor Liu Lu was put up on the country’s ‘bad artists list.’ This led to her contract termination.
But the filming of series was over before the actor’s dispute with Chinese authorities. So, the producer swapped Lu’s face with another actor’s in a Chinese web series, ChinAI reported, making it the first reported incident of face-swapping in a Chinese TV show.
In India, a day before Delhi election in February, two videos of Bharatiya Janata Party (BJP) Delhi unit President Manoj Tiwari urging citizens to vote for BJP, in english and haryanvi, were sent to 15 million voters via 5800 WhatsApp groups. The videos were deepfakes, digital media firm Vice reported.
In the altered English video, Tiwari is seen criticising his political opponent and asking citizens to vote for his party.
In the current snapshot of deepfake-related threat landscape, USA and UK nationals make up about 61% of the targets, according to Deep Trace.
The technology detection firm notes that individuals in South Korea, India, and Japan make up a significant proportion of targets. This finding correlates to their research observations that deepfake activity continues to grow in Asia.
In their 2019 report, Deep Trace identified the prominence of non-consensual deepfake pornography to have accounted for 96% of the total deepfake videos online. And the top four websites that host synthetic content received more than 134 million views on videos targeting female celebrities worldwide.