Amnesty International Slammed Over AI Protest Images

This week, international human rights watchdog Amnesty International faced backlash from photojournalists and other online critics for using AI-generated images depicting photorealistic scenes of Colombia’s 2021 protests. Although there is no shortage of photographs from the demonstrations, the advocacy group told the Guardian that it opted to use artificially edited imagery to protect the identities of protesters who may be vulnerable to state retribution.

The 2021 strike — which was incited by an unpopular tax raise and then fueled by police brutality and other forms of state violence — left at least 40 people dead and many more missing, according to official figures.

Amnesty International shared the AI images as part of a since-deleted social media campaign marking the two years since the Colombian protests, paired with disclaimers that acknowledged the use of AI. Commentators online were quick to notice errors in the fake images. For instance, one of them showed a woman wearing the tri-colored Colombian flag and being dragged off by police, a familiar still from the 2021 protests. But on social media, people pointed out that the colors in the national flag were in the wrong order, and the faces of the protesters and police officers were eerily smoothed over. Additionally, the uniforms of the officers were out-of-date.

In response to the public outcry, Amnesty International has since deleted the images from its social media channels.

The organization has not yet responded to Hyperallergic’s request for comment. In an interview with the Guardian, Director for Americas Erika Guevara Rosas said Amnesty International did not want the AI controversy to “distract from the core message in support of the victims and their calls for justice in Colombia.” 

“But we do take the criticism seriously and want to continue the engagement to ensure we understand better the implications and our role to address the ethical dilemmas posed by the use of such technology,” Rosas added.

Amnesty also directly responded to the backlash online, apologizing for the misrepresentative photos and reiterating their initial intentions.

“Our main goal was to highlight the grotesque violence by the police against people in Colombia. It is important to state that the purpose was to protect people who could be exposed. But we could choose drawings or other things,” Amnesty International tweeted. 

Some members of the photojournalism and larger arts community have also shared their frustration with the mock photos since the popularization of AI over the past year has raised questions about plagiarism and job displacement.

Molly Crabapple, a New York-based writer and artist who recently authored an open letter against the use of AI-generated art, condemned Amnesty International’s use of the tool in its campaign.  

“By using AI-generated photos of police brutality in Colombia, Amnesty International is practically begging atrocity-deniers to call them liars,” Crabapple tweeted. “Either use the work of brave photojournalists, or use actual illustrations. AI-generated photos just undermine trust in your findings.”