In recent days, the videos generated by artificial intelligence have surfaced online supposed to show dramatic scenes from the Iran-Israeli conflict, including a woman generated by the A-Rapport of a fire prison in Tehran and false images of high-rise buildings reduced to rubble in such Aviv. Other manufactured visuals represent a slaughtered Israeli military aircraft.
These clips, some who have accumulated millions of views on platforms such as X and Tiktok, are the last An increasing scheme Videos generated by AI that spread during major events.
CBS News
Researchers at the Clemson University Media Center for CMIMINISTICS told CBS NEWS that part of the content was amplified on X by a coordinated account of accounts promoting the Iranian opposition messaging – with the aim of undermining confidence in the Iranian government.
Videos manufactured
Monday, Israel made strikes on several sites In Iran, including the famous Evin prison. A few minutes after the attack, a video began to circulate on X and other social media platforms showing an explosion at the entrance. The video is grainy, in black and white and seems to be images of the security camera.
But several visual anomalies indicate that images may have been created using artificial intelligence, according to experts, including an incorrect sign above the door and inconsistencies with the explosion.
Hany Farid, professor at the University of California, Berkeley, and co-founder of the AI detection startup, Getreal Labs, told CBS News that he thought that the video had perhaps been generated by an AI empty image tool.
Farid said that the progress of recent technology has contributed to more realistic videos with easier to create and share them quickly.
“A year ago it was [that] You could make a single image that was pretty realistic, “said Farid.” Now it’s a full -fledged video with explosions, with what looks like a portable mobile device imaging. “”
The video had been published on X in the minutes following the Israeli attack on June 23 on the installation by an account which “carries brands of inauthentication”, according to researchers from criminalist media.
Iranian and Israeli officials did not comment on the authenticity of the video.
Darren Linvill, co -director of the Forensics Media Hub, told CBS News, another video, who represented a journalist generated by AI outside the prison, is the perfect example of a coordinated network using AI to broadcast false information to a wider audience.
“He does not do anything that we could not do with previous technology, he only does everything cheaper, faster and on a larger scale,” said Linvill. It’s not clear that is behind the videos, Linvill said.
Social media platform responses
When asked about Iran-Israeli videos generated by AI-AI on their platform, a spokesperson for Tiktok told CBS News that the platform did not allow disinformation Or Contents generated by Ai-Fakes Sources or crisis eventsAnd deleted some of these videos.
An X spokesperson referred CBS News to their Characteristic of community notesAnd said some of the video publications generated by AI have added community notes to help fight false information.
As for how to avoid prey to the videos created with AI, Farid said: “Stop getting your news from social media, especially on the breakdown of events like this.”
