Meta's Imagine AI was caught generating racially inaccurate images of historical figures and certain groups just weeks after Google's AI tool did the same.
According to Axios, Imagine AI supposedly generated images of the Founding Fathers of America as Black Americans and "a group of people in American colonial times" as Asian women.
The "diversity" dial on Meta's AI results was similar to Google Gemini after the latter reportedly generated images of World War II German soldiers and British monarchs as people of color.
Google has since removed the feature from its AI, promising to relaunch it in the "next couple of weeks" once the issue has been fixed.
Meta has yet to issue a statement regarding the recent findings.
Also Read: Google Gemini Pauses AI Generation of People After 'Inaccuracies' in Historical Figures
AI Warned to Deepen Racial Stereotypes, Biases
Experts and digital analysts have long warned about the potential of AI being abused to further spread racial stereotypes and biases thanks to its development process.
AI training usually involves datasets scraped across the internet, creating models based on previous stereotypes and biases prevalent online in contrast to actual depictions in real life.
These issues have been noted as leverage for certain groups to spread radical political ideas or push false narratives against minorities.
So far, far-right groups have started using AI-generated images to promote Islamophobia online, many of which depict Muslims with Middle Eastern features.
The AI posts came amid calls for a ceasefire as thousands of Palestinians, mostly Muslims, are killed every day in Israel's ongoing air strikes on Gaza.
Related Article: OpenAI's Near-Realistic AI Videos: The Dangers to Digital Information Landscape
US Regulations on AI Development Remains Elusive
Despite the apparent risks in the technology, the US government is still lagging in regulating the development and distribution of AI technology in the country.
Congress recently required AI companies to disclose information on the development process of their next AI projects but otherwise left actual guidelines up to the tech firms.
As of writing, most departments work from the president's executive order to address advancements in AI.
