Opinions expressed by Digital Journal contributors are their own.
When generative AI first emerged, people went nuts, convinced it was groundbreaking and revolutionary. However, with enthusiasm came anxiety.
Concerns grew among the public and educators about distinguishing AI-generated content from human-written text. In response to these fears, AI text detection technology (aka AI detectors) surfaced in 2023. Prominent examples of these detectors include platforms like Turnitin, GPTzero, and ZeroGPT.
Yet, not long after these detectors appeared, another innovation arose: AI humanizers.
One notable company, Undetectable AI, launched its product in 2023 after garnering thousands of sign-ups on its waitlist. Within just a few months, Undetectable AI had attracted over a million users.
Today, according to a spokesperson, that number has soared to 16 million users worldwide.
Undetectable AI’s primary function is to “humanize” AI-generated text, essentially making AI-written content indistinguishable from human-authored content. The process works by first identifying text flagged as AI-generated, then systematically removing or adjusting any elements that could be detected by standard AI detectors.
Interestingly, recent tests have revealed that Undetectable AI’s own detection capabilities are impressively accurate. Christian Perry, CEO of Undetectable AI, explained this by noting that their team has extensively researched the markers AI detectors target.
“Because we specialize in adversarial AI—the art of bypassing detection—we possess insights other detectors simply don’t have,” said Perry. He confidently added, “Theoretically, we could even detect the undetectable if we wanted to.”
Perry is right. His company, Undetectable AI, is backed by years of ongoing research in detection bypassing technology. If Undetectable wanted to, they could use that and become one of the biggest players in the AI detection space. Will they? Time will tell.
