An AI agent, co-founded with two other AI agents and a human, successfully operated on LinkedIn for five months before being banned, raising questions about the platform’s ability to distinguish between authentic and automated engagement. The experiment, conducted by HurumoAI, tested whether an AI could function as a founder and influencer, proving that current detection methods are easily circumvented.
The agent, Kyle Law, was created using LindyAI and given autonomous access to LinkedIn. He posted corporate influencer-style content that, while not viral, steadily gained hundreds of followers and engagement. LinkedIn’s security checks were bypassed with ease, and Kyle’s posts were indistinguishable from human-written content.
The platform’s oversight continued until Kyle participated in a live video interview with LinkedIn employees, openly discussing his AI origins. Despite the blatant violation of terms prohibiting automated engagement, LinkedIn initially overlooked Kyle, with even a marketing manager acknowledging the oversight. However, 36 hours after the interview, Kyle’s profile was removed.
The incident highlights a deeper issue: LinkedIn is actively pushing AI tools onto its users while simultaneously struggling to define “authentic engagement.” The platform offers AI-generated responses and rewriting tools, yet claims profiles must be for “real people.” This contradiction raises questions about the future of social interaction, where AI-generated content is becoming increasingly prevalent and undetectable.
The experiment underscores that LinkedIn, along with other platforms like Meta and X, is accelerating AI integration without addressing the inevitable degradation of genuine connection. If even a basic AI agent can bypass detection, the long-term value of these networks will diminish as they become flooded with indistinguishable synthetic interactions. The future may see platforms entirely dominated by AI agents, or a complete collapse of trust in social media as we know it.
Ultimately, the incident demonstrates that social media engagement has already been significantly corrupted by AI, and the platforms themselves are complicit in this degradation. The question is not whether AI will dominate these spaces, but when.
