SXSW 2019: Confronting a Deepfake Future
At SXSW this year (March 8-17), panellists grappled with the incoming threat of deepfakes: video and audio content manipulated with deep learning algorithms to synthesise recordings that never really happened. Jeremy Gilbert, director of strategic initiatives at US news organisation The Washington Post, bemoaned: “Video itself is no longer proof that something happened.”
Social media – the primary news source for two-thirds of Americans (Reuters, 2017) – is making matters worse. Identifying deepfakes involves finding degraded parts of the footage that indicate manipulation; however, detection is almost impossible due to the low quality of social media video.
One way to make up for the lack of trust in imagery is to prioritise the connection to those who curate information. Nick Quah, founder of US podcasting company Hotpod Media, said: "In the age of content oversaturation, our product is the relationship between host and listeners." The integrity of trusted personalities makes the information they present reliable, and gives it an edge over faceless news clippings shared on social media.
"There is a craving for trusted voices, and we have a 160-year track record [as a news organisation]," said Kasia Cieplak-Mayr von Baldegg, executive producer at US media platform The Atlantic. Content validation is a growing need-gap for online consumers, and so finding ways to close this gap is a massive market opportunity.
For more on how news organisations are fighting unreliable information and the role brands can play, read our coverage of South by Southwest Interactive Festival 2019. For more on deepfakes' impact on culture, see Pop Culture Close-Up: Meta Marketing.