Tech

Is this woman AI generated or not? The internet isn’t sure

She’s… both?

Photo of Marlon Ettinger

Marlon Ettinger

woman talking(three split)

It’s a familiar scene—you’re scrolling through Reels and suddenly a seemingly down-to-earth person appears on your screen. While they’re looking down at you as if you’re having a video call, they tell you about a tip, trick, or hack you might be interested in.

Featured Video

“So let me get this straight,” a young woman sitting in a car asks in a recently viral video. “You guys are telling me that when you’re out of the house for hours, you’re comfortable walking around with all that stanky body that’s been building up on you all day?”

Then comes the product pitch, at which point your interest is either piqued or you swipe away.

The aformentioned ad, however, was hard to swipe away from because the people posting it were claiming that it was AI despite looking shockingly real.

Advertisement

“wait. is this AI???” asked @philwinkle.

“Yes sir,” answered @beckylitv.

But while the video does seem pretty convincing, it didn’t have everybody fooled. Some posters pointed out that the woman in the video has multiple Fiverr profiles where she goes by the name Ariel M. and offers User Generated Content (UGC), product reviews, and video ads.

Advertisement

The confusion was complicated by the fact that videos of the creator talking were at least partially AI generated by a platform called Arcads.ai.

According to the terms and conditions of  Arcads.ai, which is run by a French company called FRESHR, customers can generate videos by scrolling through a catalog of pre-recorded content made by different creators in different environments, “each with its own particularities.”

“FRESHR offers it Customer the option of uploading a script … to the Platform in order to create a final video … which, on the basis of the Video Model chosen, integrates the script, in particular by making the lip-sync and cloning the voice with that of the Video Model Creator or another Creator.”

The videos are the latest frontier in user UGC, the term advertisers use for those relatable, first person videos that are constantly scrolling across our feeds.

Advertisement

According to some people in the business, the increasingly convincing applications of AI which videos like these are displaying could pose real ethical problems soon.

“Firstly, this is just v1. This technology has evolved massively in the last 3-4 months,” wrote Alex Cooper, the co-founder of ad company AdCrate on X. “By the end of the year, I predict that you won’t be able to tell the difference between human and AI on your newsfeed (at least for talking head videos).”

“Legally, it’s HUGE grey area. And it really wouldn’t surprise to see platforms/governments enforce legislation meaning you have to disclose that content is AI generated.”

Cooper also pushed back against those who argued that the technology would have a devastating impact on society.

Advertisement

“I think this will have much less of an impact than everyone thinks it will (at least for now). Ultimately, you can only generate talking head videos without product and I don’t see you being able to make footage with product anytime soon.”

That point was echoed by @beckylitv, who made the video which had people scratching their heads or doubting her.

“People are losing their minds from that AI ugc video I posted. First of all, yes it’s AI, do I look that bored to you?” she posted. “Secondly, just like with midjourney or chatgpt, you need to be skilled at prompts. This video looks real because I matched the script with the AI model. I analyzed the model and then I added text based on her mannerisms. This will still involve a lot of human skill, don’t get ahead of yourselves.”

But some people thought the original framing of the video as completely AI generated was deceptive.

Advertisement

“The ‘video looks real’ because it’s a video of real people 😄,” replied @BenjaminDEKR. “Tell people how this actually works, don’t gaslight them that it’s fully AI for clicks.  You’re being deceptive and less than honest, then attacking people when they ask valid questions.”


The internet is chaotic—but we’ll break it down for you in one daily email. Sign up for the Daily Dot’s web_crawlr newsletter here to get the best (and worst) of the internet straight into your inbox.

 
The Daily Dot