Deepfakes
Present scenario?
Could we stop for a second and think of how often we hear about artificial intelligence nowadays? We hear CEO’s, tech magazines, developers, data scientists, engineers, entrepreneurs and even journalists talk about AI. It has advanced many fields like healthcare, finance, education, agriculture, manufacturing, production, transportation and much more. If this technology is as good as it sounds then why is it making headlines about its disastrous consequences? No, it isn’t about the machines taking over the world, it’s about deepfake.
What is deepfake?
Deepfake is an AI-based technology that is used to manipulate visual and/or audio content using deep neural networks. An existing piece of content is used as a reference with the person, object or environment altered. The alteration takes place in such a way that the new content is superimposed over the old, perfectly matched so that it looks almost seamless. This creates such convincing footage that can fool anyone. You may ask what is the point here, it is that this technology is used to spread misinformation. 2019 and 2020 so far have witnessed many instances where deepfakes have managed to fool people on the internet. During this year’s Indian election campaign a major political party used deep fake to create a video criticising a State government and created lots of chaos. The spectrum is much wider in the western countries. Studies have shown that many Americans cannot tell what news is fake and what news is real. “The Pew study suggests that fake-news panic, rather than driving people to abandon ideological outlets and the fringe, maybe accelerating the process of polarization: It’s driving consumers to drop some outlets, to simply consume less information overall, and even to cut out social relationships.” It makes it harder for people to see the truth. Nowadays there is a rise in politically motivated malevolent deepfakes designed to stoke oppression, division and violence. They are being made to influence elections and are shared on various social media platforms. Another concerning revelation by a private AI firm suggested that 96% of deepfake videos were weaponised against women for revenge purposes. The technology has advanced to an extent that forged videos can be easily created using a still image as a reference. Such fake content can have a huge impact on many aspects of ordinary people’s lives.
It has been possible to alter footages for a long time with the help of highly skilled artists but deepfake technology has emerged as a gamechanger. Now with the development of this technology, anyone can make a convincing fake video and weaponize it for political or other malicious purposes. For centuries audio and video has functioned as a proof for the occurrence of an event but a time has come where we can no longer use these as evidence. The social media platforms have placed such manipulated and fabricated media as real. It can be difficult to detect a well-made deepfake, so many have gone undetected. Companies are still developing tools to detect them and are yet to establish a foolproof method. There might not be a definite solution to this problem shortly but one thing we can do is to regulate the platform through which these deepfakes are circulating and make it a safer platform.
Reference:
- Recognising fake news by ACC library services
- “why deepfakes are making headlines with its notoriety” by Sejuti Das, Analytics India Magazine
- “Deepfakes- a terror in the era of AI” by Yufei Han, Norton life lock blogs
- “Deepfakes- how do you prepare your organisation for a new kind of threat” by Koen Putman, Accenture Insights
