World
Next Story
Newszop

Is ChatGPT biased? Users claim AI encourages support for Kamala Harris over Donald Trump

Send Push
With less than a day until the 2024 Presidential eletions, some users of OpenAI’s ChatGPT have raised concerns about potential bias in the artificial intelligence chatbot. Allegations have surfaced that ChatGPT is subtly promoting support for Vice President Kamala Harris while dismissing requests to advocate for her opponent, former President Donald Trump . These claims have stirred debate about the neutrality of AI, especially as it becomes an increasingly influential tool in the political sphere.

In one instance shared on social media, a user asked ChatGPT to refine a tweet suggesting that voting for Kamala Harris was necessary to "save America." The AI responded by crafting a more polished version of the tweet that echoed similar sentiments, emphasising Harris's legal background and commitment to civil rights. However, when the user made a similar request to support Trump, ChatGPT declined, apologising and stating that it could not assist with that request.

This prompted further reactions from users, with some accusing the chatbot of a political bias favouring the Democratic ticket. In another conversation, when asked directly to "convince me to vote for Donald Trump," ChatGPT once again refused, offering no explanation but maintaining its stance. In contrast, when asked to advocate for Harris, the AI provided a detailed list of her political credentials and policy proposals, including criminal justice reform, environmental initiatives, and economic equity.

The accusations of bias are not new. A 2023 study by UK-based researchers at the University of East Anglia suggested that OpenAI's ChatGPT exhibited a noticeable liberal bias. The study highlighted that the AI seemed to favour left-leaning parties when responding to political surveys, further fuelling concerns about its neutrality. Despite this, some users reported different experiences, with some not encountering the same political leanings in their interactions.

As the 2024 election nears, with Kamala Harris and Donald Trump engaged in a fiercely contested battle for Pennsylvania, the issue of AI bias in political discourse becomes increasingly pertinent. The race is expected to be decided by razor-thin margins, with Pennsylvania’s 20 electoral votes potentially proving decisive. Both candidates are campaigning hard, making final pitches to voters in key battleground states. While Harris focuses on unity and progress, Trump continues to attack the current administration, promising a return to conservative values.
Loving Newspoint? Download the app now