AI-generated image of Oscar statuette falsely linked to LA wildfires

Days after massive wildfires broke out in Los Angeles on January 7, 2025, social media users shared an image of an Oscar statuette with a false claim that it was found in a destroyed home. The picture bears signs of digital manipulation, and an expert told AFP the image's creator confirmed it was generated using artificial intelligence (AI).

"The Oscar statuette among the remnants of the Los Angeles fire," says the Indonesian-language caption of the photo shared January 13 on X.

Image
Screenshot of the false post taken January 24, 2025

The posts emerged about a week after the deadly blazes killed at least 27, with dozens of people still missing (archived link).

A-list actors, musicians and other celebrities were among the tens of thousands of people affected by the fires, as the entertainment industry in the city screeched to a halt.

Italian actress Isabella Rossellini shared the image on Instagram on January 13. The picture also spread in online media reports and on Facebook -- including in Greek, SpanishTurkishKoreanPersian, Burmese, Chinese and French.

However, the image is AI-generated.

The InVID-WeVerify verification tool from veraai.eu, a project in which AFP is a partner, indicated there was a 95 percent chance the picture was created with AI. This detector identifies specific traces left by AI image generation software.

Image
Screenshot of an InVID analysis of the AI-generated image

A reverse image search revealed the picture was previously posted January 12 on the social media platform Reddit and titled "Symbolism" (archived link). The picture was later removed and replaced with a note saying it violated the forum's rules barring AI-generated pictures.

A watermark also appears in the image's bottom-right corner, further indicating it was fabricated using AI.

Image
Screenshot of the image shared on Reddit, with a watermark highlighted by AFP

Emmanuelle Saliba, chief investigative officer at GetReal Labs, said in a January 15 LinkedIn post that the deepfake detection company's systems determined the image "was likely to be synthesized or modified" (archived here).

Saliba told AFP in a January 22 email that she got in touch with the creator, who identified himself as a graphic designer. He told her he originally shared it in a Facebook group dedicated to AI creations and digital art and that his intention was "not to mislead people."

Saliba added that the creator told her he used Google's generative AI tool -- Gemini -- to generate the image on his phone before editing it in Adobe Photoshop.

"This is an example of how quickly an image can spread and be taken out of context," she said. "This will become more common, especially since we are moving towards a world where this type of content can easily be created on our phones."

AFP has verified several claims about AI images and the Los Angeles fires here and here.

Is there content that you would like AFP to fact-check? Get in touch.

Contact us