I opened up my Instagram feed one day and saw a video of Jake Paul “doing makeup with James Charles.” Videos like this then began to bombard my Instagram feed until I had to ask myself a question I’m sure others are currently asking themselves: is this real?

The answer was no—the video was generated by artificial intelligence (AI). And this is simply the tip of the iceberg. If you use social media, then you’ve probably seen the depths of AI usage on any platform. From Instagram AI-generated videos of Jake Paul to videos of your birthday, your bed, to President Donald Trump’s now-deleted video about developing Gaza, AI haunts every aspect of social media, and thus, every user’s algorithm. 

What’s even bigger than the depth of these videos’ content is the scale it’s produced on; an AI-generated video about a dog saying “I love you” is a part of the 96% of people that now use AI on social media. While most of these AI videos may seem harmless, that doesn’t include the AI misuse that lurks within that 71%: deepfakes.

Deepfakes are nonconsensual AI-generated videos or images of someone, and on Sept. 30, with the release of Sora 2, Open AI’s new app where users can AI-generate realistic media of anything, deepfake creation on social media skyrocketed. For example, a user created a fake Sora video about Dora getting arrested.

Bill Peebes, the head of Open AI, even reported on his X page that Sora 2 received a million downloads in fewer than five days. 

Now, users are creating AI-generated deepfakes of Jake Paul, Martin Luther King, Jr., and even screaming squirrels at a door camera with Sora 2. Users who watch these videos may not even know they are fake due to how realistic they appear. 

While, to some, these videos may initially seem funny or lighthearted, when the videos center around a real person, they are technically posting nonconsensual media of someone, violating privacy and spreading misinformation. 

These supposedly harmless videos have become hateful quickly. For instance, users have created fake movie trailers, with some making fun of 9/11 and others of people with Down syndrome. 

While the new Sora 2 generative AI fuels user’s imagination, it can fuel hatred as well. Giving users the ability to create anything they want comes at a cost, and, in this case, it’s spreading hate on social media. 

People who have personally experienced events that users now use AI to make fun of probably feel upset and hurt. Even worse, this won’t be the last time they see hatred again due to the ripple effect of using AI on social media.  

Deepfakes also serve an even darker purpose for creating nudes and sexually explicit images.

THORN, a child safety nonprofit, found from a 1,200-people survey of 13-20 year olds that one in 17 people were victims of AI deepfake nudes. Besides the already terrifying existence of deepfake nudes, the survey suggests something scarier: someone you already know could have been targeted by deepfake images on social media. 

And the effect of deepfake nudes on social media is detrimental. Deepfake nude victims are not only violated, but they have to live knowing that strangers saw unconsensual and suggestive images of them on social media. Additionally, by posting deepfake nudes, one user enables other users to follow their behavior, thus creating a ripple effect that spreads the message that violation is okay, promoting harmful values to users online. 

While social media platforms have content moderation, it often takes a while for harmful content to be deleted. Content moderation is not fast enough to stop hate, thus users need to take action.

So instead of scrolling past a problematic AI post, report it to the social media platform. By doing so, users give the platform an early opportunity to delete harmful content. Additionally, users should not take to the comment section to share their opinions about a post. Interacting with content only boosts its chance to pop up again on others and your algorithm. In this case, preventing the spread of hate and misinformation is imperative. 

Users can also post about the problems of deepfake images to help inform users about the situation. By doing so, they can create a positive ripple effect on social media that encourages users to spread information instead of hate. 

Social media is ultimately supposed to be a place that fosters positive communities; however, in this new age of AI use, communities form under a shared hate. If this continues, then there’s no telling how many more people will fall victim to deepfakes. However, if people begin to use social media for its intended purpose, then maybe there is hope for social media to become an overall positive place again.