Truth in the age of AI

As a kid(even now as well) I was quite occupied by this problem of lies and wanted a solution. Some solution where liars could be punished in real time. Some of the mythological stories I read as a kid were scornful of this idea. The story goes like this according to Claude. 

According to a 15th-century legend about Rome's Bocca della Verità (Mouth of Truth), a woman suspected of adultery was brought before the stone mask that supposedly bit off the hands of liars. Her clever lover disguised himself as a madman and ran from the crowd to embrace and kiss her in front of everyone. When she placed her hand in the mouth, she truthfully swore she had never been touched or kissed by any man except her husband and "that madman" everyone had just witnessed. She kept her hand, fooling both the lie detector and her suspicious husband.

There is also a counter story for this. 

The most famous demonstration of Solomon's wisdom came when two women—both living in the same house—came before him with a dispute. Both had given birth to sons around the same time, but one baby had died. Each woman claimed the surviving child was hers. With no witnesses and no evidence, Solomon called for a sword and declared his judgment: the baby would be cut in two, with each woman receiving half. One mother accepted this solution, saying if she couldn't have the baby, neither could the other. But the second woman cried out, "Give the baby to her, just don't kill him!" Solomon immediately declared the second woman the true mother, recognizing that a real mother would sacrifice her own claim to save her child's life. When all Israel heard the verdict, they held the king in awe, because they saw that he had wisdom from God to administer justice.

These stories demonstrate the quest for truth verification has been an everlasting pursuit of mankind. AI has made this pursuit 10x harder. Traditional methods of proofs such as pictures, videos et al can be created easily with Gen AI tools. So what is the way around it ? Obvious solutions come to mind, such as

1. Mandatory watermarks.

2. Maybe more IRL evidence,

3. Blockchain solutions et al.

But this game has always been a cat and mouse game with each out witting the other.

Our tool KYCC is an attempt at solving this by “Fighting fire with fire”.  Only AI can detect the fakeness of AI images. The trick is to highly contextualise the detection and do a comprehensive detection.

Here is our tool. https://www.dheemai.com/kycc.

Try and let us know what do you think.