22 Jun, 2024 08:59
Wikipedia’s political bias is spreading to AI – study
Wikipedia has a tendency to portray right-wing political figures in a negative light, a pattern that is feeding into AI large language models (LLMs) that harvest data from the online encyclopedia, a US-based conservative think-tank has claimed.
A report released on Thursday by the Manhattan Institute looked at evidence of political bias in English-language articles on Wikipedia, by correlating the names of Western leaders and prominent politicians with emotion-laden words.
The study found “prevailing associations of negative emotions (e.g., anger and disgust) with right-leaning public figures; and positive emotions (e.g., joy) with left-leaning public figures,” suggesting “evidence of political bias embedded in Wikipedia articles.”
“We find some of the aforementioned political associations embedded in Wikipedia articles popping up in OpenAI’s language models,” the report concluded.
The researchers noted that Wikipedia articles are likely a prominent part of OpenAI’s “secret corpus of data” used to train ChatGPT.
The report acknowledges, however, that this pattern is not universal, and is more common in articles about US political figures, while there was no evidence of Wikipedia bias in entries on British politicians or US-based think tanks.
For example, in references to recent US presidents, Donald Trump – now the Republican presidential frontrunner – was portrayed with the most negative sentiment, while Barack Obama was listed as having the most positive references.
The report concluded that Wikipedia “is not living up to its stated neutral-point-of-view policy.”
The policy referred to, which Wikipedia describes as one of the pillars of the encyclopedia, stipulates that articles must exclude personal opinions and interpretations of the editor, be based on reliable sources, and explore multiple points of view when dealing with a controversial topic.
Wikipedia has been repeatedly criticized for its supposedly biased takes on hot-button political issues, with its co-founder, Larry Sanger, saying last year that the website had become an instrument of “control” used by the US intelligence agencies to wage information warfare.
