A social media influencer campaign can appear to be performing well: tens of posts praising a product, hashtags gaining traction, and an apparent steady stream of consumers talking about a brand.
Yet some of that ostensible enthusiasm may not be as organic as it looks.
The growing use of artificial intelligence (AI) to mass-produce social media content is creating a new problem for companies buying influencer marketing without knowing that they are paying for manufactured influence and not genuine engagement.
A new investigation by the US AI giant Anthropic has lifted the lid on how the same AI tools being used to create political propaganda are repurposed for commercial marketing.
Between December 2025 and August 2026, Anthropic identified and removed an account used by an actor to mass-produce Kenyan political content through its popular Claude chatbot.
The operation was designed to make coordinated messaging appear like spontaneous grassroots commentary by Kenyans on platforms like X and Facebook.
The actor would provide a topic, talking points, and hashtags and instruct Claude to generate batches of exactly 50 posts, formatted for publication across social media.
While a large portion of the campaign focused on Kenyan politics, the same model was used for Kenyan retail brands, which Anthropic did not disclose.
‘The actor ran the identical AI workflow for Kenyan retail brands under a marketing persona, ‘SHANKI’/’Elkins Marketer,’ the report said.
The AI-generated promotional broadcast was repackaged as organic-looking tweets, with links inserted into every fifth post.
‘This fits into a commonly observed pattern in Kenya where agencies pay local influencers to manipulate narratives,’ said Anthropic.
Anthropic said the activity appeared consistent with an entirely domestic Kenyan operation.
The trend creates a new risk in Kenya’s rapidly expanding influencer economy. Companies may struggle to tell the difference between genuine creator influence and content manufactured at scale.
Influencer marketing has become big business as companies shift some of their advertising budgets away from traditional media toward digital creators with specific audiences.
Creators produce sponsored photos and videos, embed product links in posts, and sell physical and digital products directly through their pages.
Nairobi-based research and data analytics firm OdipoDev recently reported that Kenyan influencers collectively earned about Sh1.07 billion from brand-sponsored posts last year.
This creates an incentive for creators and agencies to produce more content, reach more people, and demonstrate engagement to brands.
AI can also lower the cost an influencer incurs – instead of writing many variations of a promotional message, an AI system can produce hundreds of posts in seconds, tailored to different audiences and social media platforms.
But while the technology can help a creator brainstorm ideas, improve grammar, edit video, generate design concepts, or adapt a genuine campaign to different platforms, it becomes problematic when it starts manufacturing the appearance of independent consumer interest.
Brands pay influencers and agencies because they believe the creators have audiences that trust their product recommendations. If the campaign is built around AI-generated accounts, comments, or posts designed to look like independent opinions, the company risks paying for an audience and influence that do not really exist.
This is similar to the political tactic of astroturfing, where coordinated messaging is made to appear as if it is coming from ordinary members of the public.
Anthropic classified the Kenyan operation as Category One on its Breakout Scale because the activity remained isolated within a network of fake accounts and local influencers on a single platform and did not reach or influence real people.
‘While a social media site usually sees an operation once its content is already circulating, we may see it on Claude while the operation is still being built,’ Anthropic said.
‘Actors use AI to plan their campaign, choose their targets, and write the material. Those types of tasks produce signals that our systems are trained to detect, which often lets us disrupt an operation before it gets off the ground.’
British newspaper The Guardian has reported cases of content creators in Europe producing AI influencer material for corporate clients, who asked them to sign non-disclosure agreements preventing them from discussing their work.
In Kenya, new rules proposed under the Artificial Intelligence Bill, 2026, would require AI systems generating or manipulating images, voice or likeness to clearly label the output as AI-generated.
This would create increased scrutiny as companies seek to establish how their campaigns are produced, whether creators are using automated tools, whether images and testimonials depict real people, and whether engagement around a campaign comes from genuine users.