ARTIFICIAL intelligence can now produce surprisingly realistic fake images – and it is putting us all at risk of being tricked, cyber security experts say.
Scammers may use fake images to trick you into handing over money or information.
1
For example, a romance scammer could create catfish accounts using AI-generated images – and even convincing videos with fake voices.
And savvy trolls can use these “deepfake” videos or AI images to spread false information News.
The problem is that it is now immensely difficult to detect whether a photo is AI-generated or not.
“Like chatbots, AI Images are increasingly difficult to distinguish from man-made images,” said cybersecurity expert Paul Bischoff, speaking to The US Sun.
“There may not be any indication that you are seeing an AI-generated image.”
Bischoff, consumer privacy advocate at Comparitech, added: “Instead, you should consider the source and whether you trust it.”
Many applications can generate highly realistic images in just a few seconds.
There are many free apps that do this – and even more powerful premium versions.
This means that even low-budget scammers can pull off surprisingly professional deepfake scams.
SAVE YOURSELF
Many images are now so realistic that it is effectively impossible to differentiate them from reality.
And this will only get worse over time, meaning you can’t rely on trying to detect poorly rendered AI hands or eyes.
“This becomes increasingly difficult,” Dr. Martin J. Kraemer, security awareness advocate at KnowBe4, told The US Sun.
“At this point, I just wouldn’t trust the image anymore.
“It is safe to assume that any image could be fake.”
If something doesn’t feel right, stop.
Dr. Martin J. Kraemer
But there are ways to help detect fake images.
Kraemar said there are three clues you look for to see if an AI deepfake is being used to trick you.
Tip #1: “If possible, apply common sense. If the image is used as a means to emotionally manipulate you, then you will likely be subject to a social engineering attack.”
Tip #2: “You may also be able to search for the image on the internet if it has been previously indexed by Google. But be careful, attackers also manipulate Google search results.”
Tip #3: “If something doesn’t look right, stop, check the source, look for other information, and try to track down the image or any contextual information.”
Deepfakes – what are they and how do they work?
Here’s what you need to know…
- Deepfakes are fake videos of people that look perfectly real
- They are made using computers to generate convincing representations of events that never happened
- This often involves swapping one person’s face with another’s or making them say whatever you want.
- The process begins by feeding an AI with hundreds or even thousands of photos of the victim
- A machine learning algorithm swaps certain parts frame by frame until it generates a realistic but fake photo or video
- In a famous deepfake clip, comedian Jordan Peele created a realistic video of Barack Obama in which the former president called Donald Trump an “imbecile.”
- In another, Will Smith’s face is pasted onto the character Neo in the action film The Matrix. Smith turned down the role to star in the failed film Wild Wild West, while the Matrix role went to Keanu Reeves
This story originally appeared on The-sun.com read the full story
