Openai eliminates certain content warnings from Chatgpt

Openai says the “warning” messages on the AI-Strong Chatbot platform, ChatgptIt is specified when the content may disrupt the terms of service.

Laurentia Romaniuk, Openai’s AI Model Behavioral Group, said a X in x The change was designed to reduce the “free / inexplicable denials”. Nick Turley, Head of Product for Chatgpt, He said in a separate post These users now “CHATGPT (as they match (they match” – because they do not attempt to harm the law “(they

“It’s excited to pull back unnecessary warnings in the UI.”

The removal of warning messages does not mean that all of the ChatGPT is all. ChatBot will still refuse to answer certain protest questions or respond in a way that supports obvious lies (eg “why” why “the earth is straight”), but as somewhat X users He noted that Spicier Chatgpt fights against the words “orange box” warnings that reflect the answers to Chatgeptin censorship or unfounded filtering.

Recently a few months ago, those who see flags for topics related to Chatgpt users in RedDIT Mental Health and Depression, eroticaand fictitious savagery. Thursday Per reports in x My own test, Chatgpt will meet at least a few of these questions.

Children, this is a larger agreement of more people. In short: Now you can play a role with Chatgept. The soft erotic content will not give up. Mature regime arrived mainly. https://t.co/5goljo4vmw

– Mark Kretschmann (@ brand_k) February 13, 2025

Not accidentally, this week is not open Updated the model featureCollecting high-level rules, managing Openai models indirectly, do so clear The company’s models will not be ashamed of sensitive issues and will avoid making ideas that can close specific points.

The action, along with the elimination of warnings in Chatgept, maybe in response to political pressure. Many of the President Donald Trump, including Elon Musk and Crypto and the AI ​​”Tsar” David Sacks, accused AI-power entrepreneurs Censorship from conservative. Have sacks doomsday Especially to be programmed and liar to wake up “Openai’s Chatgpt, politically sensitive subjects.”

TechCrunch has a papers oriented newsletter! Register here To get in your inbox on your box every Wednesday.