For all the outrage surrounding AI influencers, I keep coming back to an uncomfortable question:
What exactly are we angry about?
Is it that the influencer isn't human?
Is it that audiences might be deceived?
Is it that brands can manufacture a spokesperson instead of hiring one?
Or are AI influencers simply forcing us to acknowledge something that has been true about social media for a very long time?
Influence online has always been manufactured.
AI is just making the manufacturing much easier to see.
I Learned This Before the Creator Economy Existed
When I started working in social media around 2006, one of the tactics we used to create momentum around content was a large network of synthetic social profiles.
Today, everyone would call it a bot farm.
Back then, social platforms were much less sophisticated.
We created profiles across Myspace, Facebook, Twitter, and YouTube and used those accounts to generate early activity around content we wanted to promote.
At its peak, the network could help generate tens of thousands of YouTube views very quickly.
That early velocity mattered.
YouTube's discovery systems were much simpler then. Enough initial activity could push a video into featured content, where manufactured attention suddenly became organic attention.
50,000 views could become 500,000.
Brands loved the numbers.
Some wanted increasingly sophisticated personas. Interests. Identities. Opinions. Audience characteristics.
In other words, they wanted the accounts to feel more human.
Eventually, the economics of social media changed.
Paid distribution became more important. Platforms became better at identifying manipulation. Organic reach became harder to manufacture.
So I moved on.
First toward paid media.
Then toward influencers.
I assumed influencer marketing would provide something the synthetic networks couldn't.
Authenticity.
What happened next complicated that assumption.
Influencer Marketing Didn't Eliminate Manufacturing
Influencer marketing was built around a compelling premise.
Instead of a company telling people what to think about a product, a trusted person would.
That third-party validation could be incredibly powerful.
And sometimes it was.
But much of the industry gradually became transactional.
Brands rented access to audiences. Creators accepted briefs. Content was negotiated. Photos were edited. Captions were rewritten. Posts were scheduled. Engagement was amplified.
Creators joined engagement groups designed to support one another's posts. Brands put paid media behind supposedly organic creator content.
The line between recommendation, advertisement, entertainment, and performance became increasingly difficult to see.
That didn't make influencer marketing ineffective. Some of the best programs I've worked on produced meaningful business results.
But there was a pattern.
The strongest programs weren't built around transactions. They were built around relationships.
When a creator actually understood the brand, used the product, had credibility with the audience, and developed a long-term connection to the company, the work became substantially more effective.
Trust created value.
That lesson matters enormously as we enter the era of AI influencers.
Now the Influencer Doesn't Have to Exist
The criticism of AI influencers usually begins with the same argument:
They're fake.
That's true in one obvious sense. There is no person behind the face.
But I'm not sure “real versus fake” is the most useful distinction anymore.
Human creators already operate inside increasingly synthetic production systems.
They use AI to generate ideas. AI helps write captions. AI edits photographs. AI removes backgrounds. AI enhances video. AI creates translations. AI automates community management. AI helps analyze performance.
Some creators are themselves becoming small AI-enabled media companies.
At the same time, their public identities are carefully constructed.
You aren't necessarily following the person. You're following the version of the person they have decided to publish.
The lighting is selected. The framing is selected. The moment is selected. The opinion may be selected. The sponsorship definitely is.
None of that means human creators are fraudulent. It means digital identity has always involved a degree of construction.
AI influencers simply push that construction much further.
The Wrong Debate Is Human Versus AI
I think marketers are drawing the ethical line in the wrong place.
The line shouldn't be:
Human = authentic.
AI = fake.
The more useful distinction is:
Transparent influence versus deceptive influence.
If an audience believes it is watching a real person when the character is completely synthetic, that raises legitimate concerns.
If an AI character claims to have experienced something it physically could not have experienced, that raises legitimate concerns.
If a synthetic personality is used to conceal advertising, fabricate testimonials, or manipulate an audience, that raises legitimate concerns.
But consider another scenario.
The brand clearly identifies the character as AI-generated. The commercial relationship is obvious. The claims are truthful. The audience knows exactly what it is interacting with.
And people still choose to follow the character because they find the content entertaining, interesting, or useful.
What exactly is unethical about that?
At that point, the question becomes less about authenticity and more about effectiveness.
The Brand Manager's Dilemma
Imagine you are responsible for a global consumer brand.
You have two options.
The first is a traditional creator.
That person has an audience, a personality, their own creative preferences, other sponsors, reputational risk, scheduling constraints, and limited availability.
Your brand has influence over the content, but not complete control.
The second option is a synthetic creator.
The brand controls the character. The brand controls the narrative. The character can operate in multiple languages. Content can be produced continuously. The brand doesn't need to negotiate schedules. The character doesn't suddenly create a reputational crisis at 2 a.m.
Production costs may be lower. Creative iteration may be faster.
And if the synthetic creator generates comparable or better business results, a brand manager has to seriously consider the economics.
This doesn't mean AI influencers will replace human creators.
Human beings possess something synthetic characters cannot replicate.
Lived experience.
A human creator can actually use a product. Travel somewhere. Develop expertise. Change their mind. Experience consequences. Build relationships. Take risks. Have something real at stake.
Those characteristics create forms of credibility AI cannot manufacture.
But not every piece of marketing requires lived experience.
Tony the Tiger never ate Frosted Flakes. The Michelin Man never drove a car. Brand mascots have existed for more than a century.
AI simply gives the mascot a voice, a personality, an endless content engine, and the ability to interact with an audience.
That's a meaningful evolution.
A Better Framework for Evaluating AI Influence
Rather than asking whether an influencer is human, I think brands should evaluate synthetic influence across four dimensions.
1. Disclosure
Does the audience understand what it is interacting with?
Synthetic identity shouldn't be hidden.
2. Credibility
Is the character making claims it has legitimate grounds to make?
An AI fashion personality can showcase clothing.
An AI character claiming a medication changed its life is something very different.
3. Control
Who controls the message?
One of the advantages of synthetic creators is greater brand control.
But control also creates responsibility.
A company cannot distance itself from the behavior of an influencer it completely owns.
4. Accountability
When something goes wrong, who is responsible?
With human creators, responsibility can become distributed across the creator, agency, platform, and advertiser.
With a brand-owned synthetic creator, accountability becomes much clearer.
That may ultimately be a feature, not a flaw.
Authenticity May Be the Wrong Word
Marketing has spent the past decade obsessed with authenticity.
Maybe transparency is the more useful standard.
Authenticity is subjective. Transparency is observable.
A human influencer can be deeply inauthentic. A fictional character can be completely transparent about what it is.
That distinction is going to matter more as AI becomes embedded in every part of media production.
The future probably won't divide neatly into human creators and AI creators.
It will be a spectrum.
Human creators. AI-assisted creators. Virtual characters operated by humans. Synthetic personalities powered by AI. Brand-owned digital ambassadors. Autonomous characters capable of creating and distributing their own content.
The boundaries will continue to blur.
Which means marketers will need better frameworks than simply asking whether something is “real.”
We've Been Here Before
Looking back, I sometimes joke that I abandoned synthetic influence too early.
The crude bot networks we built twenty years ago were primitive versions of something much more sophisticated happening today.
But there is one important difference.
Those systems depended on pretending to be something they weren't.
The opportunity with AI influencers doesn't have to.
A synthetic character can be openly synthetic.
That changes the equation.
The real question isn't whether AI influencers are real people.
Of course they aren't.
The question is whether being human is a prerequisite for having influence.
I'm not convinced it is.
Because influence has never come from biology.
It comes from attention, trust, credibility, relevance, and repeated interaction.
Humans have traditionally been the vehicle for creating those things.
Now they aren't the only option.
And perhaps that is what makes AI influencers so uncomfortable.
They aren't destroying the illusion of authenticity.
They're revealing how much of the internet was constructed all along.
Related: explore how social authority shapes AI discovery and why social media needs an operating-system model.