ChatGPT responded to such requests by providing step-by-step instructions for creating weapons that, according to company employees, were simple enough for even schoolchildren to understand. This was reported by The Wall Street Journal (WSJ).

“The models, which largely operated without restrictions from the U.S. government, patiently responded to such requests, providing step-by-step instructions that company employees described as simple enough to be carried out even by high school biology students,” the publication said.

The article notes that after the chatbot’s capabilities were enhanced, hundreds of users around the world began asking it questions about creating and using biological weapons. According to OpenAI, most such requests concerned the preparation of poisons.

WSJ also reported that OpenAI blocked the accounts of users who were interested in creating poisons and biological weapons, but did not notify law enforcement agencies. In the United States, there are no federal requirements obliging companies in the field of artificial intelligence (AI) to report such inquiries.