ChatGPT under fire again: WSJ publishes controversial report

The influential American publication The Wall Street Journal has published controversial information regarding artificial intelligence safety. The report states that some users attempted to obtain dangerous information about biological weapons and poisons through ChatGPT, and experts noted that in some cases, the system provided detailed answers.

According to the article, after the OpenAI models were updated, various users sent queries on dangerous topics related to biology and chemistry. Experts pointed out that some of the answers were technically accurate.

At the same time, OpenAI stated that accounts associated with dangerous queries are being blocked, models are being continuously retrained, and safety restrictions are being tightened. Company representatives noted that strict filters have been introduced in current systems to prevent the provision of such information.

According to WSJ, there is currently no unified federal law in the US requiring AI companies to mandatory report such queries or automatically forward them to law enforcement agencies.

The publication also wrote that some users searching for dangerous information appealed to other major chatbots, but no precise information was provided on what answers they received.

Companies such as OpenAI, Anthropic, and Google emphasize that they are cooperating with governments and independent security experts to prevent the misuse of artificial intelligence for harmful purposes.

According to experts, due to the rapid development of artificial intelligence technologies, safety issues are becoming one of the main global discussion topics. Therefore, major tech companies report that they are working on regularly strengthening the protection mechanisms of their models.