Zoom has become the first major platform to adopt a detection tool developed by OpenAI to distinguish between human and AI-generated content. The move comes as generative AI tools grow more sophisticated, raising concerns about misinformation and fraud. OpenAI’s AI Classifier will be integrated into Zoom’s services to flag content suspected of being artificially generated.
Sam Altman’s company introduced the classifier earlier this year as part of its broader efforts to address misuse of AI technology. The tool analyzes text for patterns typical of machine-generated output, such as unnatural phrasing or repetitive structures. Zoom will use it to verify the authenticity of user-submitted content, particularly in professional and educational settings where accuracy is critical.
The collaboration marks the first public deployment of OpenAI’s detection system outside controlled testing environments. Zoom’s decision follows internal evaluations that confirmed the classifier’s reliability in identifying AI-generated text with a low error rate. Company officials stated the integration aims to build user trust by providing transparency about content origins.
OpenAI has not disclosed plans to expand the classifier to other platforms beyond Zoom at this stage. However, the company has indicated ongoing discussions with potential partners. The tool remains in active development, with periodic updates to improve detection accuracy as AI models evolve.
Experts warn that while detection tools help, they are not foolproof. Some AI-generated content can evade classifiers, especially when heavily edited by humans. Zoom has emphasized that the system will supplement, not replace, human moderation and fact-checking processes.
Source: digi.no