Technology

Fake Instagram Stars: What Happens When AI Creates Influencers

Two AI-generated Instagram influencers faked attending a major movie premiere, sparking backlash and raising questions about authenticity, competition with real creators, and how social media platform

Martin HollowayPublished 2w ago5 min readBased on 8 sources
Reading level
Fake Instagram Stars: What Happens When AI Creates Influencers

Fake Instagram Stars: What Happens When AI Creates Influencers

Two AI-generated characters named Santos Walker and Caleb Ellis posted photos claiming to have attended the red carpet premiere of The Devil Wears Prada 2. The studio that made the film had nothing to do with it. The photos were completely fake, created by AI. When the hoax became public in April, it sparked outrage across social media. The incident has raised a straightforward question: what happens when AI-generated people start pretending to be real influencers.

These are not the only synthetic influencers out there. A character named Jae Young Joon, created by a Canadian named Luc Thierry, has accumulated over 320,000 Instagram followers. Another AI person called Romeo DeSouza, designed to appear Dutch-Brazilian, has 56,000 followers. The creators of these fake accounts are linking up with each other, sharing techniques and strategies for building audiences and earning money from their fake personas.

What Social Media Platforms Are Doing

When users reported the fake red carpet photos, Instagram's parent company Meta began looking into complaints about AI-generated accounts. But it is not yet clear what rules Meta is actually enforcing, or how strictly. Many of these synthetic accounts continue posting without any clear warning to followers that they are looking at content made by AI, not a real person.

People online have flagged what they see as a growing wave of fake profiles. Many of these AI accounts show attractive men, often claiming to be gay, and sometimes mentioning recovery or sobriety. A lot of them reuse photos and ideas from real influencers without permission. In effect, a whole shadow industry of fake people is competing for attention and brand deals alongside actual human creators.

Real influencers are understandably worried. An AI person can post new content every single day without getting tired, traveling, or dealing with personal problems. A human creator has to manage their real life. That difference in workload could eventually mean fewer brand partnerships for people who are actually doing the creative work.

Why People Are Following Fake Influencers

It is interesting to look at who is actually following these AI accounts. Jae Young Joon's followers are mostly women, even though the account appears designed to appeal to gay men. This mismatch suggests that either the algorithm is amplifying the account in unexpected ways, or the fake influencer's content resonates across different groups in ways its creator did not predict.

Some of these accounts are more honest than others. Romeo DeSouza's profile clearly states it is an AI creation. Many others do not. It is not yet clear whether the creators made this choice themselves or whether Instagram is pushing them to be transparent.

Behind the scenes, the people making these AI influencers are staying in touch, sharing what works and what does not. This is no longer just hobbyists experimenting with AI. It is starting to look like a real business.

The Money and the Broader Concern

These synthetic accounts are making real money. The exact figures are not public, but it likely comes from the same places that real influencers earn income: sponsored posts, links that earn commissions when people buy products, and subscription services. The advantage for AI creators is that generating content costs almost nothing once the AI is set up.

This pattern has played out before in technology. When YouTube started paying people for videos, it created the same kind of tension between traditional media companies and new creators trying to make money on their own. But there is an important difference now: AI can remove humans from the process entirely. Before, platforms were making it easier for regular people to compete with big media. Now, they are making it possible to replace people with AI.

Some in the gay community have raised concerns about the bodies shown in these AI accounts. Santos and Caleb were actually mocked online for looking unrealistic — their muscles were drawn too large, too bulky. But other AI influencers are carefully designed to look like idealized versions of what people wish they looked like. For followers who believe they are looking at real people, that can hurt their self-image. The creator of Jae Young Joon has acknowledged these worries and seemed to understand why human influencers are upset.

What This Means Beyond Instagram

This is not limited to lifestyle and entertainment. Thousands of AI-generated accounts claiming to support political candidates have appeared on social media in recent months. These fake personas are spreading into politics alongside fashion and lifestyle content.

The scale and sophistication of what is happening suggests something important: AI is now cheap and easy enough that you do not need technical expertise to create a convincing fake influencer. The barrier to entry keeps getting lower. The potential money keeps growing. That combination is creating conditions where these experiments turn into actual businesses.

When fake people can pose in front of real movie premieres and fool thousands of followers, it raises hard questions for the companies running these platforms. Their rules were written for content made by humans on accounts run by humans. AI-generated accounts are starting to slip through the gaps in those rules. The red carpet hoax might end up being the moment when platforms finally decide they need stricter policies about what counts as authentic content and who gets to use that label.

The bigger question is less about two specific fake influencers and more about what comes next. As AI gets better at creating realistic-looking content, and more people figure out how to use it, platforms and regulators will have to catch up with technology they were not built to handle.