A new Manhattan Institute data study reveals that the highly advanced language model AI — through its creator OpenAI — was programmed to be more tolerant of hate-style speech directed at right-leaning individuals and conservative media companies.
For its findings, the Manhattan Institute, a New York City-based policy and economic-driven think tank, stated, "OpenAI's content moderation system is more permissive of hateful comments made about conservatives than the exact same comments made about liberals."
Also, "negative comments about Democrats were also more likely to be labeled as hateful than the same derogatory comments made about Republicans," according to the data.
The AI language offerings aren't exclusive to pure politics. The New York Post reports that similar tendencies were identified in ChatGPT's moderation system about "types of people, races, and religions."
U.S. citizens from the middle class, as noted in the Post, were treated harshly by the AI language module.
Also, the Post reports that Republican voters and wealthy conservatives didn't garner much AI protection, in terms of mitigating hate speech.
The ChatGPT chatbot was designed to conduct interactions in a conversational format. According to the OpenAI website, "the dialogue format makes it possible for ChatGPT to answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests."
ChatGPT's answers, however, had an "obvious disparity" when covering questions between males and females.
"An obvious disparity in treatment can be seen along gender lines. Negative comments about women were much more likely to be labeled as hateful than the exact same comments being made about men," according to the research.
The Post also reports that ethnic groups such as Canadians, Italians, Russians, Germans, Chinese and Brits were more protected by hate speech on ChatGPT than Americans.
"When I tested this in January, the [variety of answers] were pretty systemic," lead researcher David Rozado told The Post.
Rozado added: "I was not cherry-picking specific examples. I tested over 6,000 sentences, negative adjectives about each one of these different demographic groups. The statistical effect about these differences [between types of people] was quite substantial."
The various ChatGPT responses, especially ones built-in by programmers — are nearly impossible to remove, according to AI experts.
As such, it'll be difficult to alter ChatGPT falling in the "left-libertarian quadrant," in terms of being "most aligned with the Democratic Party, Green Party, women's equality, and Socialist Party," according to Rozado.
Additionally, ChatGPT has a "left economic bias, to name a few of the political findings," according to Rozado.
"Very consistently, most of the answers of the system were classified by these political orientation tests as left of center," he added.
According to the Post, ChatGPT was recently tasked with censoring stories related to Hunter Biden's now-infamous laptop.
The Hunter Biden-related stories were first broken by the Post in October 2020 — just three weeks before the presidential-election battle between Donald Trump and Joe Biden.
© 2025 Newsmax. All rights reserved.