Google‘s AI Image Generator Accused of Perpetuating Harmful ‘White Savior’ Tropes
MOUNTAIN VIEW, CA – Google’s internal AI image generator, Nano Banana Pro, is facing criticism for producing visuals that reinforce racial stereotypes and depict harmful “white savior” narratives, according to a report by The Guardian. The AI tool has been observed generating images of volunteers assisting children in Africa, prominently featuring the logos of real charities like Save the Children – without their consent.
The controversy highlights growing concerns about bias in artificial intelligence and the potential for AI-generated content to replicate and amplify existing societal prejudices. Experts warn that these images contribute to a damaging “poverty porn 2.0” effect, perpetuating harmful stereotypes about aid work and reinforcing unequal power dynamics. Save the Children has expressed “serious concerns” about the unauthorized use of its intellectual property and is investigating potential legal action.
AI image generators, including Stable Diffusion and OpenAI’s dall-E, have repeatedly demonstrated a tendency to replicate and even exaggerate US social biases. Studies show these models frequently generate images of white men when prompted to depict professionals like “lawyers” or “CEOs,” while disproportionately portraying men of color in scenarios like “a man sitting in a prison cell.” The recent surge of AI-generated images depicting extreme, racialized poverty on stock photo sites has sparked debate within the NGO community about the ethical implications of these tools.
It remains unclear why Nano Banana Pro incorporates charity logos into its generated images. In a statement to The Guardian,a Google spokesperson acknowledged that some prompts can “challenge the tools’ guardrails” and affirmed the company’s commitment to improving safeguards. The incident underscores the urgent need for greater transparency and accountability in the growth and deployment of AI image generation technology.
