In December 2022, Edward Tian, a Princeton University student, developed a prototype for GPTZero, a ChatGPT-detection tool, from his dorm room. As the popularity of ChatGPT surged, Tian grew concerned about the potential difficulty in identifying AI-generated content disguised as human-authored articles or essays. He spent his winter break building GPTZero and then secured seed funding during spring break.
Tian expressed concern about ChatGPT’s rapid rise in a Forbes interview, stating, “It was suddenly thrust upon the world, and there’s lots of potential for misuse.”
GPTZero, which has gained 1.2 million users over the past five months, announced on Monday that it raised $3.5 million in seed funding from early-stage venture firms Uncork Capital, Neo, Altman Capital, and Stability AI’s CEO Emad Mostaque. Reuters CEO Tom Glocer and former New York Times CEO Mark Thompson also participated in the round.
Capable of detecting text from various AI models, including Google’s LaMDa (Bard), Facebook’s LLaMa, and OpenAI’s GPT-3 and GPT-4, GPTZero partnered with educational organizations like Canvas and K16 Solutions to obtain training data for its AI detection tool. Educators found the tool helpful in identifying student essays generated using ChatGPT.
The startup plans to offer its services to enterprise customers in publishing, social media, and trust and safety sectors, and envisions the government as a potential client, which could utilize the AI detection tool to inform policy. The company currently has 3,000 paid subscribers for its $9.99 per month plan.
GPTZero will soon launch a browser plugin called Origin to evaluate the accuracy and origins of facts in published text. Through Origin, the startup aims to promote media literacy in the AI era by fact-checking AI-generated content and verifying citations. Tian emphasized, “This is about more than catching students misusing the tools — it’s not even about detecting the AI. It’s about preserving what’s human.”