AI-generated fake news is one of the biggest dangers to our society.
People use AI to influence brand reputation, the stock market, political elections, and more in these 4 ways:
A deepfake refers to any realistic video, image, or audio clip made using AI that mimics a real person. Deepfakes can make it look like a person has done or said something they wouldn’t in reality, thus causing outrage, confusion, or fear.
Example: Taylor Swift Endorsing Donald Trump
During the 2024 presidential race, Trump shared on Truth Social AI deepfakes of Taylor Swift and her fans appearing to proudly endorse him.
Trump fans delighted in the idea that a popular musician had finally backed him. Meanwhile, Democrats scrambled to beg her to quickly correct the public with her actual opinion.
Over the last few years, the quality of AI-generated images has gone from laughably bad to very realistic. Now, people use tools like DALL-E 3, Midjourney, and Stable Diffusion regularly.
People question the truthfulness of images less than written text. In the chaos of social media, AI images are often accepted as true and cause the rapid spread of fake news.
Example: Explosion near the Pentagon
In 2023, social media users spread this AI-generated image of an explosion near the Pentagon like wildfire on social media. This resulted in a dip in the stock market as panic caused share sales.
Closer inspection revealed clear signs that AI had generated the building and gates. Arlington PD quickly responded and debunked the image on X. However, a verified account impersonated Bloomberg and reposted the image, causing further panic.
Social media bots are automated accounts that imitate humans for a wide range of nefarious purposes and can spread fake news by:
Example: Russian Internet Research Agency (IRA) US Election Interference
Russian IRA operatives use AI bots to promote polarising propaganda on social media sites. They create fake accounts to target specific demographics of voters to increase the spread of sensationalist and conspiratorial fake news. Operatives often create accounts across the political spectrum from the same device to exacerbate social divisions.
AI can automatically create vast amounts of content, like news articles and YouTube videos, to farm passive advertising revenue. This content is not designed to be accurate and often has no human supervision whatsoever. It is only designed to get attention.
Example: Fake Halloween Parade in Dublin
An automated fake news site released an article about a fictitious Halloween Parade that it claimed was taking place in Dublin. This site was based out of Pakistan. Hundreds of people lined the streets in preparation for it after word had got around, only to discover that there was no event.
Learn about 5 AI risks to your business
We’re approved contractors to US Federal Government CAGE code 88BP5 “Social Simulator Inc, The”
