Zürcher Nachrichten - 'Misrepresent reality': AI-altered shooting image surfaces in US Senate

Social media platforms are awash with graphic footage from the moment US agents shot and killed 37-year-old intensive care nurse Alex Pretti in Minneapolis, which sparked nationwide outrage.

One frame from the grainy footage was digitally altered using artificial intelligence, AI experts told AFP.

The manipulated image, which purports to show Pretti surrounded by officers as one points a gun at his head, spread rapidly across Instagram, Facebook, X, and Threads.

It contained several digital distortions, including a headless agent.

"I am on the Senate floor to condemn the killing of US citizens at the hands of federal immigration officers," Senator Dick Durbin, a Democrat from Illinois, wrote on X Thursday, sharing a video of his speech in which he displayed the AI-enhanced image.

"And to demand the Trump Administration take accountability for its actions."

In comments beneath his post, several X users demanded an apology from the senator for promoting the manipulated image. On Friday, Durbin's office acknowledged the mistake.

"Our office used a photo on the Senate floor that had been widely circulated online. Staff didn't realize until after the fact that the image had been slightly edited and regret that this mistake occurred," the senator's spokesperson told AFP.

- 'Advancing an agenda' -

The gaffe underscores how lifelike AI visuals -- even those containing glaring errors -- are seeping into everyday discourse, sowing confusion during breaking news events and influencing political debate at the highest levels.

The AI-enhanced image also led some social media users to falsely claim the object in Pretti's right hand was a weapon, but analysis of the verified footage showed he was holding a phone.

That analysis contradicted claims by officials in President Donald Trump's administration that Pretti posed a threat to officers.

Disinformation watchdog NewsGuard said the use of AI tools to enhance details of witness footage can lead to fabrications that "misrepresent reality, in service to advancing an agenda."

"AI tools are increasingly being used on social media to 'enhance' unclear images during breaking news events," NewsGuard said in a report.

"AI 'enhancements' can invent faces, weapons, and other critical details that were never visible in original footage -- or in real life."

The trend underscores a new digital reality in which fake images -- created or distorted using artificial intelligence tools -- often go viral on social media in the immediate aftermath of major news events such as shootings.

"Even subtle changes to the appearance of a person can alter the reception of an image to be more or less favorable," Walter Scheirer, from the University of Notre Dame, told AFP, referring to the distorted image presented at the US Senate.

"In the recent past, creating lifelike visuals took some effort. However now, with AI, this can be done instantly, making such content available to politicians on command."

On Friday, the Trump administration charged a prominent journalist Don Lemon and others with civil rights crimes over coverage of immigration protests in Minneapolis, as the president branded Pretti an "agitator."

Pretti's killing marked the second fatal shooting of a Minneapolis protester this month by federal agents.

Earlier this month, AI deepfakes flooded online platforms following the killing of another protester -- 37-year-old Renee Nicole Good.

AFP found dozens of posts across social media, in which users shared AI-generated images purporting to "unmask" the agent who shot her.

Some X users used AI chatbot Grok to digitally undress an old photo of Good.

L.Muratori--NZN