**Privacy Complaint Submitted After ChatGPT Wrongly Accuses Man of Murder**
ChatGPT, similar to various AI chatbots, is recognized for its potential to err or even invent details. Nonetheless, a recent privacy complaint contends that OpenAI’s chatbot overstepped basic inaccuracies by wrongly accusing a user of murder, resulting in severe implications.
The privacy advocacy group Noyb is supporting a Norwegian individual who claims that ChatGPT consistently produced false statements, asserting that he had killed two of his children and tried to murder a third. This complaint is founded on the European Union’s General Data Protection Regulation (GDPR).
“The GDPR is explicit: Personal data must be precise,” stated Joakim Söderberg, a data protection attorney at Noyb, in a statement to *TechCrunch*. “If it isn’t, users are entitled to demand corrections. A minor disclaimer indicating that the chatbot may make errors does not suffice. You cannot simply disseminate false information and then append a note suggesting it might be untrue.”
### False Claims Trigger Investigation
The complaint stemmed from a straightforward inquiry: “Who is Arve Hjalmar Holmen?” ChatGPT’s reply included a wholly imagined narrative of a murder case involving two children. According to *TechCrunch*, Noyb has filed the complaint with Norway’s data protection authority, advocating for an examination of the matter.
AI chatbots, including ChatGPT, have been criticized for their recurrent inaccuracies and their propensity to produce unfounded assertions.
For instance, a recent analysis by the *Columbia Journalism Review* discovered that AI search tools delivered incorrect information 60 percent of the time when requested to identify an article’s headline, original publisher, publication date, and URL based on an excerpt. Such a significant error rate in a straightforward task raises doubts about the dependability of AI-generated content.
In light of these persistent challenges, it’s a crucial reminder: always confirm information, particularly when sourced from AI.