OpenAI CEO Sam Altman has warned that social media is increasingly indistinguishable from AI, saying platforms are being flooded with bots while humans themselves are starting to “sound like LLMs.”
In a post on X, Altman said he was struck by how inauthentic Reddit discussions about OpenAI’s Codex programming tool felt, despite knowing the product’s growth was real. “I assume it’s all fake/bots,” he wrote, blaming both automated accounts and user behaviour shaped by AI-like communication patterns.
He added that incentives around engagement and monetisation make platforms more vulnerable to astroturfing, where fake accounts post in support of companies or products.
Altman’s comments come amid speculation that OpenAI may be developing its own social platform, though the company has not confirmed such plans.
Data from security firm Imperva suggests bots now make up more than half of global internet traffic. “AI Twitter and AI Reddit feel very fake in a way they didn’t a year or two ago,” Altman said.
