Google bans deepfake projects on its Colab platform

Google has banned deepfake projects on its Colab platform, which offers computing resources for Python code. Deepfakes are synthetically-generated media used to deceive and manipulate, with the potential for serious consequences like political sabotage and personal harm. The ban reflects a growing concern over the rising prevalence of deepfakes online.

Developers cannot use resources made available via Google’s Colaboratory platform (Colab) to generate deepfakes. Colab provides access to computing resources, enabling users to run Python code directly through the browser; the platform is typically used for machine learning and data analysis projects. 

The ban seems to have been added to Colab’s terms of use earlier this month, but was only recently spotted. 
The battle against deepfakes is one worth waging. Deepfakes are synthetically-generated media (images, video) in which AI is used to make it appear that someone did or said something that they did not. Deepfakes can be misused in different contexts: to discredit opponents in political campaigns, to cause damage to someone’s personal reputational (for instance, by making people appear in pornographic videos they were not part of), and even to escalate situations that can lead to violent conflict. Statistics indicate that between 2019 and 2021, the number of deepfakes available online increased from 14,000 to 145,000.