Fake news
We see and hear news every day and trust that the information provided is accurate. That belief may soon end.
Artificial intelligence (AI) software is now being developed that can produce fake video footage of public figures using recordings of their own voices. Using as little as one minute of user-generated content (data), it can reproduce a particular person’s voice. The developer of this software demonstrated the results by using the voices of Bill Clinton, George Bush and Barack Obama in a computer-generated conversation.
Once a person’s voice has been reproduced, a fake video can be created by processing hundreds of videos of the person’s face. Video footage of politicians are often used, as there is so much data available online.
Law professor John Silverman commented that, as humans we tend to believe what we see, and the increased number of tools to make fake media that is unrecognizable from real media is going to prove a major challenge in the future.
Discuss the claim that companies who develop software that can create fake videos of politicians should be accountable for the fake videos posted by users of their software on social media platforms.
Essay preview
Companies that develop AI software capable of generating fake videos of politicians should be accountable to some extent, but not fully liable for every fake video posted by users on social media. The strongest position is a shared-accountability model: developers are responsible for the foreseeable risks created by th