In the last two years people have created videos, called deep fakes, that use artificial intelligence to alter faces or to make people say things they never did. Google is now trying to help fight a battle against these videos, and has released a database of 3,000 deepfakes
It is using videos of actors created by using a variety of publicly available tools to alter their faces. The plan is to help researchers to build the tools they need to detect and remove “harmful” fake videos.
It is feared that these videos may be used to promote false conspiracy theories and propaganda.
So how does this work in practice?
Deepfake technology starts of by using video and audio clips of real people, and then uses artificial-intelligence techniques to modify them in various ways such as putting words in their mouths or transposing their heads on the body of someone else carrying out some distateful activity.
Google said:
“Since the field is moving quickly, we’ll add to this dataset as deepfake technology evolves over time and we’ll continue to work with partners in this space.
“We firmly believe in supporting a thriving research community around mitigating potential harms from misuses of synthetic media.
Meanwhile, Facebook has announced it has set up a $10m (£8.1m) fund to find better ways to detect deepfakes. Indeed they are very aware of the dangers of deepfakes since their own chief executive, Mark Zuckerberg, has been a victim of the technology, which involved him appearing to state in a manipulated video that a secretive organisation was responsible for the success of the Facebook social network.
