The Stream

Would you be fooled by a deepfake?

Artificial intelligence is getting better at creating ultrarealistic fake videos.

This is the first show in a week-long series highlighting themes from the 2019 SXSW Conference and Festivals.

Video has long been considered a source of hard evidence in determining the truth. But how will trust in video change at a time when artificial intelligence is making it easier to create fake audiovisuals?

Synthetic videos, also known as deepfakes, are created using AI-powered programs that generate faked data based on real data sets. The technology has been used in making face swapping videos, fake porn, and fake audio clips using the voices of real people.

If misused, deepfakes could potentially have catastrophic implications. Fake videos could be made to foment social discord, sway an election, used for blackmail or for government propaganda.

As the ability to make deepfakes becomes more advanced and accessible, the technology may reach a point where even machines won’t be able to tell the difference between what’s real and fake. In this episode, we’ll discuss the future of fake videos and their potential impact on society.

On this episode of The Stream, we speak with:
Sam Gregory @SamGregory
Programme Director, WITNESS

Tarun Wadhwa @twadhwa
Founder & CEO, Day One Insights

Tim Hwang @timhwang
Director, Harvard-MIT’s Ethics and Governance of AI Initiative

Read more:
Don’t believe your eyes: Deepfake videos are coming to fool us all – SecurityIntelligence
Artificial intelligence, deepfakes, and the uncertain future of truth – Brookings

What do you think? Record a video or leave a comment below.