Google Quietly Bans Deepfake Creation from Colab Platform
Google updates its Colaboratory terms to explicitly prohibit users from generating deepfakes. The quiet policy change targets the use of free cloud computing power for creating non-consensual synthetic media.
Google officially prohibits users from creating deepfakes on its Colaboratory platform, a service that provides free browser-based access to powerful cloud computing resources. The company quietly updates its FAQ section to include deepfake generation on the list of disallowed activities, directly targeting tools like DeepFaceLab that rely on Google's hardware to bypass the need for expensive personal graphics cards.
This policy shift comes shortly after an investigation highlights how DeepFaceLab frequently directs users to deepfake pornography sites for software tutorials. A popular fork of the software, known as DFL-Colab, allows people to leverage Google's servers specifically to avoid buying high-end equipment, making the platform a crucial hub for non-consensual deepfake creation.
A Google spokesperson states that the ban results from regular reviews of abusive patterns that contradict the company's AI principles. While the platform currently blocks all deepfake generation, Google expresses a desire to eventually distinguish between malicious deepfakes and benign applications as its detection methods improve over time.