AI Influencers Are Coming
Brands are starting to deploy marketing dollars on AI-generated influencers. What does that mean for the future of influencer market and social media?
If one trend dominated social media marketing in 2025, it was the rise of microinfluencers.
Instead of paying mega-creators upfront, brands started recruiting thousands of smaller creators (1–50k followers), gave them unique affiliate links, and paid them sales commissions. Brands essentially turned microinfluencers into a performance channel. One of the best examples is Comfrt, which sold $700 million of weighted hoodies to Gen Z with the help of small TikTok creators.
I think a similar but different trend will dominate 2026: the rise of AI-generated microinfluencers. Brands are going to use armies of AI avatars to promote their products on social media.
Don’t believe me? Check out these videos.
If these videos popped up on your feed, you’d have no idea they were AI.
AI models are finally nearing the point where you could generate a fully realistic avatar promoting any product you want, and saying anything you want.
In that world, brands would be crazy not to use AI influencers.
Social media algorithms reward high output of threshold-quality content. If you’re a creator, you don’t know what content will go viral. So it makes sense to produce as much solid content as possible, and some of it will catch fire.
If you’re a brand and want 500 microinfluencers talking about your product on TikTok, you could either pay real humans $1K per video for $500K total. Or you could spend $5 in AI credits per video for $2.5K total.
The AI-generated option isn’t just cheaper. It’s faster to deploy. It doesn’t require a partnerships lead DMing back and forth with creators. And most importantly, the system can self-optimize. Once a brand recognizes what content is performing best, it can double down on those content forms.
X and YouTube are full of agencies explaining how to generate armies of AI influencers.
As more brands leverage AI influencers, what are the downstream consequences for the $30 billion influencer marketing industry? I’d hypothesize three things.
First, I don’t think A-list influencers are going anywhere. The word influencer connotes… influence. Kim Kardashian and Taylor Swift will not lose their influence. But our social feeds continue to move away from “who you follow” and towards “your algorithm.” People are getting more accustomed to seeing creators they don’t recognize. Brands are spending more affiliate dollars on microinfluencers. And I do believe AI avatars will displace some of that real-human microinfluencer spend.
Hot take: I also think A-list influencers will use AI versions of themselves to avoid travel and time on set for brand deals. Here’s one early example: James Harden recently became the brand ambassador of an online casino, and the announcement video was fully AI-generated.
Second, I think short-term value will accrue to the brands who lean into AI influencers, and to the agencies who set them up. I recently met two early-twenties founders building one of these agencies. They have 1,000 iPhones sitting on cooling racks, each automated to generate a different influencer, “warm up” the accounts with non-branded posts, and then post promotional content for brands. Setting up this physical infrastructure and software workflows takes technical skill and scrappiness. That’s why many of these agencies are run by college kids. But if you can drive low CPMs for brands, you will make money.
Third, I think long-term value accrual depends on how social media platforms handle the influx of AI influencers. Social media platforms have a long history of cracking down on the latest CPM hack. In social’s early days, brands used armies of bots to inflate likes, so Facebook built bot detection systems to remove those likes. More recently, brands started clipping previously-viral content, so Instagram tweaked the algorithm to prioritize original content.
The platforms have already tried cracking down on AI Slop. Last year TikTok started requiring labeling of AI-generated content, and banned AI content posing as real-life content. Last week New York State passed legislation requiring brands to disclose AI content. However, as AI content converges on 100% realism, I question whether it’ll be possible to identify what’s real and what’s fake. If not, it becomes a race to the bottom. Every brand will exploit the lower CPMs of AI influencers. Our feeds will get flooded with AI influencers. And social media will start to suck even more.
Recently I was at a happy hour and met the head of product at a major social media platform. He said there’s a real chance that social feeds get flooded by AI content, people stop using the apps, and we see the death of social media.
Maybe that’s a pessimistic view of the future.
Or maybe that’s the optimistic view.