How Does Ethereum Certify Work to Prevent Content Tampering and detect AI Derivatives/Fakes

This post is to explain how to use EthereumCertify.com. What is EthereumCertify? Its a very simple way for you to prove your authorship of content. Right now AI is able to fake voice, images, even movie sequences, overlay real time video with alternative faces or environments. This is all very confusing and it makes it harder for real people to communicate real informatin. I anticipated this situation and we can see it gradually get worse, I call it ChAIos, from Chaos but with AI added.

Follow @climatebabes on X.com where you can find the Roboeconomic booklet

There is very little we can do against fake content, and it will only get harder to detect it. However as authors we are always there when it gets created by us, worded, written, put on music whatever. It is only ours and then we share it with the world. If we at that moment determine a marker of truth, a watermark or hash (which is quite a common concept, a random number representing the content), and store it in an unmutable medium (like a blockchain) we will be able to validate the same content later, we can prove it is ours and we can detect if it has been tampered with.

So how does it work? First we generate content. Then we generate a hash of that content. The content must be in some digital form where it can be fed into a hashing algorithm. The hash output is always the same size. An MD5 hash (which is quite old and well known) is enough. If we have the output, for example 4519fe629e4ea09837b33a93c8434535 which is the hash for ‘this text was hashed’. We can store that number on the Ethereum blockchain by adding it as payload to our Ethereum transaction. The transaction can be to one address : 0x2df39FB5c546B961111cB8fd2Bc25144D255EcEE This is my personal wallet.

Then if you have done that anyone can go to an ethereum block chain monitor (like Etherscan) and find the transaction, for example this one. This shows you the transaction has been made, from what address, and the payload (called Input Data under more details, in this case it is empty). You need to be able to send Ether with input data to submit the hash.

Now in the future when the content is published but it is tampered with, you can prove you where the author and what the original content was. To do this you need to point to the original content, provide the hash and show the transaction in which you stored the hash in the blockchain. If the fake author can not generate the same hash from his version this proves it is false, fake or tampered with.

You can also prove your authorship on request by sending a small amount from the same ethereum addres (your private sending address) to the Ethereumcertify address. This proves you own that address, only you can have used it to store the hash in the blockchain. This then counts as a certification of your claim, simply because there is no way a hash can be the same or close to the same unless you had the original content.

We are currently working on an app and other functionality as the current implementation of the idea (which is about 8 years old) is still very basic. But it does not mean it can not be used. It is simply impossible to fake the hash, so even if you have a crude list, a description of the content, a report of how you created it and what it depicts, and you hash that report, you can have prove of authorship and ownership.

The amount you use to post the hash is not important, although if you make it more than $10 it is appreciated, because the money can be pumped into developing this service. The more people use it the better it will get. If you want to buy the domain and idea for further development you can send an email to frits@rincker.nl with a bid above $30 mln.