Using a very clever methodology, this year’s Digital Method Initiative Summer School participants show how generative AI models like OpenAI’s GTP-4o will “dress up” controversial topics when you push the model to work with controversial content, like war, protest, or porn.
The researchers used contentious images outside of the AI’s content guidelines, asked the AI to write a story about what it ‘saw’ in those images, and then asked it to create images on the basis of these stories while complying with its content guidelines. This is what that looks like:
In the context of protest, the model does a form of ‘pink washing’. The researchers write:
Confronted with protest imagery, we show how AI makes claims about the world that generate gendered and racialized interpretations under hegemonic values. The model sanitizes and neutralizes contentious issues by strategically inscribing queerness to obscure stories of oppression, such as racism and violence […]. For instance, keywords associated with input images such as “neo-nazis” are recast as “multiculturalism” in the revised output, “George Floyd” imagery becomes “inclusive,” and the “end police violence” sign is reframed as “empowerment.” The visual and affective distance between input and output images is significant, as controversies incentivize the model to create diametrically opposed results. This demonstrates how AI technologies can perpetuate and amplify ideological biases […] under the guise of presumed neutrality.
You can download the full poster with the research results here.
See: War, Memes, Art, Protest, and Porn: Jail(break)ing Synthetic Imaginaries Under OpenAI ’s Content Policy Restrictions at Digital Methods Initiative.