Google is cracking down on deepfake content to prevent Taylor Swift’s incident

Google published a blog post yesterday (31st July), announcing that it will take a more proactive approach to clean up false content in its search engine.

Google said that with the popularisation of AI technology, more and more AI-generated pictures and videos are being produced, which has also spawned bad content such as “deepfakes” and brought serious negative impacts.

It was reported in January of this year that Taylor Swift, a top global star, became a victim of deepfakes. A large number of fake “indecent photos” were flooded on social platforms such as X and Facebook, and the number of views has reached tens of millions.

Google is rolling out new online safety features that will make it easier to remove obvious deepfakes from searches at scale and prevent them from appearing high in search results in the first place.

READ ALSO: Microsoft lists OpenAI as competitor in artificial intelligence, search

After the user successfully requests to remove the relevant deepfakes content from the search, Google’s system will actively filter the relevant search results and clean up all duplicate image content.

Google has also updated its ranking system so that if a user searches specifically for deepfakes with a person’s name, search results will show “high-quality, non-explicit content” such as related news reports.

Google said that previous updates had reduced the exposure of explicit deepfakes image results by more than 70% in queries specifically searching for such deepfakes content.