Google halts Gemini AI's image generation amid historical accuracy concerns

Gemini AI image generation: Google has temporarily suspended its Gemini Artificial Intelligence (AI) chatbot's capability to generate images of people following concerns over historical inaccuracies. This decision comes in the wake of recent scrutiny after users reported discrepancies in the images generated by the AI tool.

The tech giant announced the pause after facing backlash for images produced by Gemini, which depicted racially diverse characters in settings typically dominated by white individuals. Specifically, users noted instances where historical figures like US senators from the 1800s and Nazi-era soldiers were portrayed with a level of diversity that did not accurately reflect historical records.

Addressing the issue, Google stated, “We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon.”

The Verge conducted tests on several Gemini queries, including a request for 'a US senator from the 1800s.' The results returned images that depicted what seemed to be Black and Native American women. This discrepancy highlighted a significant issue, as the first female senator in the United States was a white woman in 1922. Consequently, Gemini's AI-generated images inadvertently erased the historical context of race and gender discrimination.

Before this decision, users had expressed concerns over whether Google's AI model was overcompensating for racial biases, given the stark disparities observed in the generated images.

Responding to the controversy, Google affirmed its commitment to rectifying the inaccuracies, acknowledging the need for immediate improvement in Gemini's depiction of historical figures. Despite the missteps, the company underscored the importance of Gemini's ability to generate a wide range of diverse images, a feature that caters to users worldwide.

Formerly known as the Bard chatbot, Gemini represents Google's latest foray into the realm of artificial intelligence. Launched officially on December 6, the Gemini AI utilises the Imagen 2 model, allowing users to generate high-quality images based on text prompts without the need for external Optical Character Recognition (OCR) tools.

As Google works to address the concerns surrounding Gemini's image generation capabilities, the tech community awaits the re-release of an improved version that aligns more closely with historical accuracy while maintaining the tool's versatility and accessibility to users worldwide.