OpenAI Secretly Turns Off AI Detection Tool

Open AI has decommissioned its AI Classifier owing to poor accuracy, crashing educators’ hopes.

In January, OpenAI, an artificial intelligence powerhouse backed by Microsoft, publicized a tool meant to save the world, especially in education. In this case, it would aid teachers and professors in detecting if AI tools, for instance, ChatGPT, have been utilized to create content. The tool is nonfunctional six months later since it failed to meet its expectations.

Last week, OpenAI, ChatGPT’s creator, silently shut down AI Classifier, its AI detection tool, owing to its low accuracy rate. This explanation was not in a new declaration. Instead, it was included in a note added to a blog post that initially publicized this tool.

OpenAI Withdraws AI Classifier’s Link

AI Classifier’s link no longer exists. OpenAI wrote that it was working to integrate feedback and was also exploring more practical source techniques for text. Besides, it revealed its commitment to developing and deploying apparatuses to enable users to establish whether visual or audio information is generated using artificial intelligence.

Nearly every day, new tools permitting the utilization of increasingly advanced artificial intelligence are released, creating a cottage industry of AI detectors.

AI Classifier Proofs Ineffective to Detect AI Footprint in Challenge Test

OpenAI publicized the introduction of AI Classifier, revealing that it would differentiate text written by AI and humans. Nevertheless, the company also revealed that the tool was not totally dependable.

In this case, assessments on a ‘challenge test’ of English texts properly established 26 percent of text written using AI as ‘probable AI-written.’ Further, it wrongly considered text written by a human as AI-generated 9 percent of the time.

According to OpenAI, AI Classifier’s restrictions entail unreliability in texts having less than 1000 characters and improper consideration of text written by a person as written using an AI Classifier founded on neural platforms performing ineffectively beyond their training data.

Education Sector at Loss of Terminated AI Detector

An example of a sector with a significant interest in accurately detecting artificial intelligence in education. Since the introduction of ChatGPT in November, instructors have raised concerns regarding learners’ utilization of the chatbot to write essays.

OpenAI acknowledged that identifying text written using AI is a major point of discussion among instructors. Besides, it is critical to identify the effects and limits of text generated using this technology in the classroom.

The move to terminate the platform hyped to facilitate the educators proof of originality in school tasks is a big blow to the pursuit of ethical AI. The secretive termination raises doubt about OpenAI’s decision to revive the project in the future. 

Editorial credit: Vitor Miranda / Shutterstock.com