As artificial intelligence becomes more prominently used in our society, so does the rise of deepfakes.
And this has become worrying. According to HRGrapevine, ‘Microsoft’s president, Brad Smith, said his biggest concern around artificial intelligence is deepfakes’.
So, what exactly are they? And why are they becoming more widely talked about in the news? This blog will explain what the term means, how to spot the technology, and the latest deepfake news.
What are deepfakes?
According to WhatIs.com, a deepfake is ‘a type of artificial intelligence used to create convincing images, audio and video hoaxes.’
The greatest danger of deepfakes is their ability to spread misinformation from sources that look trustworthy.
Deepfake videos are created in two ways- either using an original video and changing what the person does or says or swapping one person’s face onto a video of another individual.
Although deepfake technology can be used for positive reasons, such as for film productions and creative campaign uses, there is a danger that this technology could be used to scam people by having what appears to be a celebrity or a person endorsing something in a video which they have never been a part of.
How to spot deepfake technology:
- Inconsistencies in the video– look for poor lip syncing, jerky movements, unusual blinking or lack of, and shifts in skin tones or lighting.
- Check facial features– skin may appear too perfect, or there may be shadows in places that you would not expect; typically around the eyes and eyebrows.
- Glasses– if the person depicted is wearing glasses, is there any glare on them? Or too much? Deepfakes fail to represent the natural lighting of a video or image.
- Facial hair– this can be added on or removed from a person’s face and can fail to look natural, making it an area to be able to spot a deepfake.
Latest deepfake news
Deepfakes are becoming more prevalent in the news lately, for positive and negative reasons. Here is a roundup of some of the latest news that has caught our attention.
Security Issues
As reported by Sky News, Britain’s cyber security agency has warned that advances in artificial intelligence could harm the integrity of the UK’s next general election.
The agency warned ‘about the prospect of increasingly realistic, deepfake videos and other forms of disinformation designed to influence voter preferences.’
The UK’s use of paper voting does make it harder for AI to interfere with voters, so the concern lies more with deepfake videos and images being used on online and social media platforms to spread misinformation about campaigns quickly.
To counteract this, the UK has set up a Defending Democracy Taskforce and a Joint Election Security Preparedness unit.
Celebrity fakes
As reported by BBC News, Indian actress Rashmika Mandanna was recently targeted with a deepfake video, which placed her face on top of an existing video which went viral.
The video was originally posted on Instagram by a British-Indian woman named Zara Patel and featured her entering a lift. This had then been altered with Mandanna’s face placed on top, making it look like she was the woman in the video.
Journalist Abhishek Kumar was the first one to report on the news and highlighted the video on his X profile, labelling it a deepfake.
This is an example of how easy it can be to manipulate footage to fool a wide range of users. The hope for society is that there will be more sanctions on this kind of behaviour, as it is scary and upsetting for the person targeted and to the people confused by the video.
Consumer PR Campaign Creativity
To finish on a more positive note, there can be a way to use the technology creatively, as well as informing users of the risk of the technology.
As reported by Marketing Dive, Steak-umm, a US food brand, ‘devised a campaign that uses deepfakes to highlight the danger the technology presents, not only to celebrities and politicians, but also ordinary consumers.’
They did this by creating a public service announcement style video where they set up a fake focus group of real vegans who are given vegan cheesesteaks and filmed their opinions of the product.
On the same day, a team behind the scenes used AI technology to alter their videos and opinions and display them as meat eaters, which they showed back to the group. As you can imagine the consumers’ reactions were not positive at first, but once they were told the campaign was actually about AI and deepfake technology, they were shocked at the results. One consumer said it made them rethink AI, as they thought something like this could only happen to celebrities.
This clever consumer PR campaign not only gets people talking about the brand for their food products, but also for the extra work they do at highlighting issues, in this case the battle of misinformation.
Summary
Overall, AI technology is continuingly growing, and with that comes the use of deepfake technology. This doesn’t have to be all negative, but it will require stricter laws and regulations to come into play in order to stop misinformation being spread and stop peoples’ likenesses being used without their consent.
If you want to find out more AI news, why not check out Escapade’s blog post on The Future of AI!
Need the help of a consumer PR agency? Please get in touch by visiting our Contact page, we would love to hear from you!
Sources
https://www.techtarget.com/whatis/definition/deepfake
https://www.bbc.com/news/world-asia-india-67305557
Blog Cover Photo: ArtemisDiana/istockphoto.com
