January 9 News, January 7Tibet6.8 magnitude in Tingri CountytsunamiThe first time I saw this, it touched the hearts of hundreds of millions of netizens. Online circulation of a "child buried map", the picture is heartbreaking, triggering a large number of forwarding and comments. Many netizens expressed heartache, and blessing, cheer. However, after verification, the picture is not a real earthquake picture, but a false picture synthesized by AI tools.
It is understood that the image originated from a self-published author on November 18, 2024 AI-generated videosThe author has labeled the video "generated by AI". A closer look at the image reveals many traces of AI generation, such as unnatural colors, lighting, movement, unreasonable movement of rocks, and even a character with six fingers. 1AI notes that the original author has now deleted the video.
According to CCTV news reports, during the earthquake, similar fake pictures and videos were spread in large numbers, with most of the images showing buildings collapsing, children being homeless, and mothers in the rubble using their bodies to hold pressure for their children. In this regard, the Tibet Internet Reporting Center, Netcom Rikaze and other relevant departments have carried out rumors, pointing out that these pictures and videos are all generated from AI, old pictures or video splicing, and calling on the public not to gullibly spread.
Many netizens are worried that this false information will deceive netizens who do not understand AI technology, and even be used by lawless elements. Lawyers pointed out that the reckless use of AI technology to generate images of severe earthquake damage or even fabricate rumors, which is seriously inconsistent with the facts and triggers widespread concern and even panic in society, is an act of disrupting public order and will face legal sanctions.
In addition, some experts are concerned that the prompt "suspected to be generated by AI technology" appears about one hour after the AI-generated pictures and videos of many online platforms are sent out, indicating that there is a delay in the platform's auditing and other problems.
CCTV News also gives ways to screen AI images as follows:
- Observe the details: many AI-generated images and videos may have stiff character expressions, blurred edges, unnatural light shadows, abnormal or distorted number of details such as fingers, etc.
- Checking the source: On some platforms and software, taking a screenshot and "recognizing the image" will allow you to find out if it is from an old image that has been published.
- Multiple verification: Do not rely on a single source of information, but obtain information from multiple authoritative sources for comparison. For some hot events, official media reports can be retrieved to verify the authenticity of the information.
- Stay alert: carefully analyze the logic of the information; if there are obvious gaps, contradictions or inconsistencies with common sense, it is likely to be false information.
- Share with caution: Don't forward information randomly until you are sure it is true and reliable, so you don't inadvertently become a promoter of rumor spreading.