Get the latest tech news
This Week in AI: It’s shockingly easy to make a Kamala Harris deepfake
In this edition of This Week in AI, TechCrunch's AI newsletter, we look at how easy it is to create a political deepfake.
Some high-profile examples from the past year include a bot network on X targeting U.S. federal elections and a voicemail deepfake of President Joe Biden discouraging New Hampshire residents from voting. The researchers say that the image- and text-analyzing models they tested failed to ignore pop-ups 86% of the time, and — as a result — were 47% less likely to complete tasks. Meta announced yesterday that it’s working with partners to make its Llama “open” AI models available for defense applications.
Or read this on TechCrunch
