False information is rampant, and traceability of images requires solutions
In an era of information explosion, can you believe that every photo is real? Whether everyone has free access to photo editing software or generative AI tools, it has become more difficult to "tell a story through pictures".
In particular, the spread of fake news and false information has become a global problem, and it is common for image data to be tampered with. A picture that has been edited or paired with wrong text is enough to mislead the public and influence the trend. Faced with this challenge, how to trace the source of images has also become a challenge faced by media and digital content creators.
Numbers Protocol from Taiwan uses blockchain technology to make pictures no longer just a supplement to news text, but has the function of self-certification, revealing information such as the photographer, shooting time, location and even the original image, making every picture All have clear origins that can be traced.
Numbers Protocol uses blockchain to record image information
Take Numbers Protocol's cooperation with Reuters during the 2020 US election as an example. Numbers Protocol specially integrated the reporter's photography equipment with the company's Capture system. At the moment the shutter is pressed, all the data of the picture, including the shooting time, Location, camera information, etc. will be uploaded to the chain to ensure that they cannot be tampered with.
In other words, no matter how this picture is later disseminated or edited, its original source can be traced through the data on the chain.
In addition to Reuters, Numbers Protocol has also cooperated with Rolling Stone magazine to investigate war crimes committed by Serbian volunteers and recorded the collected images through Numbers Protocol.
Numbers Protocol co-founder Yan Shiji explained that when journalists or photographers use this technology to shoot images, their works no longer rely solely on "media reputation" to prove authenticity, but on the source of the images that cannot be tampered with through the blockchain. record. For news media, this not only improves the authenticity of the content, but also enhances the trust of the audience.
Not only for the 2020 US election, but also for this year’s Taiwan presidential election, Numbers Protocol also cooperated with public television, TaiwanPlus, Sanli TV and even freelance journalists to put picture messages on the chain.
Yan Shiji said that in addition to media, ordinary creators can also download the Capture Cam application on their mobile phones to convert photos or videos into verifiable digital assets.
Yan Shiji, co-founder of Numbers Protocol. Photography by Cai Renyi
The Importance of “Identifying Sources” in the AI Era
Although everyone is aware of the seriousness of fake news. However, Yan Jiji admitted that many media outlets cannot afford additional technology expenditures due to financial pressure. Even so, Numbers Protocol has not given up on deepening cooperation with the media and continues to expand to more international markets, such as cooperating with media in Indonesia, India and other places.
In addition to fake news issues, with the popularity of AI image generation technology, Numbers Protocol will also focus on improving the transparency and traceability of AI-generated images.
Numbers Protocol launched the Capture Eye function this year, allowing website owners to easily embed image or video-related information into the website through simple HTML embed codes. Website visitors can directly click on the ICON on the image to view and verify the source of the content.
The Capture Eye function also has an AI-generated identification function. From the small icon, the public can tell whether the effect was generated by AI or shot by a human. In addition, the Capture Dashboard function also allows users to note whether the image was generated by AI when uploading it, and indicate which AI model was used.
At the same time, Numbers Protocol has developed an "Allow AI Mining" toggle button. When users upload images, they can choose whether to allow AI companies to use the images as training data.
Numbers Protocol strives to provide adequate AI recognition capabilities. However, Yan Shiji emphasized that "authenticity and falsehood" are difficult to define. Digital platforms cannot guarantee that all uploaded content is absolutely authentic, but they can ensure that the content is complete and cannot be tampered with. In other words, users can judge the authenticity of the content based on complete information instead of blindly trusting the information provided by the platform.
More coverage
Technology giants bring StopNCII to combat DeepFake. How to use "digital fingerprint" to delete victim content?
Google plans to launch AI generated icon function! How does the "C2PA technology" behind it solve the proliferation of Deepfake content?
