Gemini is known, but old problems persist

Gemini is known, but old problems persist. Google temporarily disabled Gemini’s human image generation feature after the AI model produced racially inaccurate depictions of historical figures, including Nazi leaders and American Founding Fathers. These erroneous images—such as depicting Adolf Hitler as a Black man or showing only white individuals as U.S. founders—sparked global outrage and reignited debates about bias in artificial intelligence.

Visual Inaccuracies and Public Backlash

Google acknowledged that the flawed outputs reflected systemic biases embedded in its training data. Historical figures of African, Asian, or Indigenous descent were consistently omitted or misrepresented, reinforcing a distorted, Eurocentric narrative of history. Social media platforms erupted with hashtags like #GeminiBias and #AIHistoryFail, as users shared screenshots of the offensive outputs. Academics and AI ethicists warned that these weren’t mere technical glitches—they were manifestations of how AI systems reproduce and amplify societal inequities, often without accountability.

A Temporary Fix, Not a Permanent Solution

Google’s response was to suspend the feature entirely. While this halted immediate damage, it did not address the root cause: biased training datasets and insufficient ethical oversight. Even when users explicitly requested diverse historical representations, Gemini continued to default to stereotypical or homogenous portrayals. Experts argue that simply turning off a feature is not innovation—it’s avoidance. Google claims it is retraining its models with more inclusive data and collaborating with external ethics boards, but timelines remain vague. Meanwhile, other Gemini functions—text generation, Q&A, and non-human image creation—remain fully operational, leaving users to wonder: if the AI can’t accurately depict history, can it be trusted to interpret it?

Gemini is hailed as an award-winning AI system, yet this incident exposes a deeper truth: advanced technology without moral integrity is dangerous. The challenge isn’t just building smarter AI—it’s building fairer AI. Gemini is known, but old problems persist. And this isn’t just Google’s problem—it’s the entire AI industry’s reckoning.