Tragic realism: how to regulate Deepfakes in Colombia?

Abstract

Recent technological advances in artificial intelligence and machine learning have made it possi-ble to create fake videos that look authentic. As these videos—known as “DeepFakes”— leave few traces of manipulation, they can be used to spread false statements and alter elections in a more sophisticated way than both online disinformation and misinformation. Considering that online disinformation and misinformation were used to capture the electorate’s decision-making in the Colombian peace referendum, deepfakes constitute a risk to Colombia’s future elections. This ar-ticle explores how other jurisdictions have responded to the emergence of deepfakes to evaluate three regulatory alternatives based on a regulatory impact analysis approach: leave the regulation unaltered; making social networks accountable for their content; and criminalizing the production and distribution of deepfakes. We conclude that, despite the criminalization of deepfakes being an alternative that could better achieve regulatory purposes in Colombia, its enforcement could also represent a threat to freedom of speech and to democracy.