Blog Post

Prmagazine > News > News > ChatGPT reportedly accused innocent man of murdering his children
ChatGPT reportedly accused innocent man of murdering his children

ChatGPT reportedly accused innocent man of murdering his children

More than two years have passed since Chatgpt exploded on the world stage, and Openai has made progress In many ways, there are still many obstacles. One of the biggest problems: Hallucinationor represent false information as facts. Now, Austrian advocacy group Noyb has submitted The second complaint against Openai For such hallucinations, it was reported that a specific instance was named, and Chatgpt reportedly said that the Norwegians were a murderer.

To do things, somehow, and worse, when the man asked Chatgpt what he knew about him, he was reportedly sentenced to 21 years in prison for killing two children and trying to murder his third child. The hallucinations are also filled with real information, including the children he has, the gender and the name of his hometown.

Noyb claims that this response made Openai violate GDPR. “GDPR is clear. Personal data must be accurate. If not, the user has the right to change it to reflect the truth,” Joakim Söderberg, a data protection lawyer at Noyb statement. “Show Chatgpt users a tiny disclaimer that a chatbot can obviously make a mistake inadequate. You can’t just spread false information and end up adding a small disclaimer that says everything you’re saying may not be true..”

Other notable examples of Chatgpt hallucinations include blaming a person Fraud and misappropriation of public fundscourt reporter Children abuse With a law professor sexual harassmentsuch as multiple publications reported.

NOYB’s first complaint About the hallucination Openai, April 2024focusing on the date of birth of public figures (so not murder, but still inaccurate). Openai rejected the complainant’s request to delete or update his date of birth, claiming that the request cannot change the information in the system and simply block its use on certain prompts. Chatgpt answered in a disclaimer that it “can make mistakes.”

Yes, there is a motto that everyone makes mistakes, which is why they put the eraser on the pencil. But when it comes to a very popular AI-powered chatbot, does that logic really work? We’ll see if and how Openai responds to Noyb’s latest complaints.

Source link

Leave a comment

Your email address will not be published. Required fields are marked *

star360feedback