In my studies and research, I recently went down the rabbit hole of AI generated images and their effects on legal proceedings. This is an important topic that is becoming more common, especially as the use or should I say, misuse, of AI grows. Now, do not get me wrong I do see good in AI, but that good is mixed with a huge bag of bad! It was the same for any technology that humanity has developed. For example, we crafted tools such as pitchforks to assist in farming and harvesting crops. But evil people will use such a tool wickedly and harm others. That is just one example! AI, as a tool, is no different. It can be a great benefit to humanity, but it can also be detrimental to our well-being.
Why do I write this? Well, as a prior law enforcement officer, who handled sensitive cases and material, I came across this form of argumentation from time to time. A suspect was arrested and charged with possession of suspected illicit material. When the defense began preparing its case it was often alleged that the material was not real people or events but had been generated using some form of computer system. Whether it was Photoshop or something similar, that argument has been around since the beginning of computer usage in the criminal field.
So, with that knowledge I grew curious as to how this latest iteration of such a criminal defense was being handled when it came to AI. I came across the group c2pa.org. This group is led by such tech giants as Adobe, Google, Intel, Microsoft and many others and it seeks to “create open and interoperable technical standards for establishing digital content provenance and authenticity.” This is a much-needed process in the world of crime, and I was thankful that such a group exists!
They utilize content credentials as a standard for creating tamper-resistant metadata, which can be used across a wide variety of content types. This includes images, audio, video and even documents. According to some sources, there are over 400 quintillion bytes a day being created! So, being able to analyze and evaluate digital content is essential.
As you can see above, even the highest office in the land is posting fake pics! I have no idea why the President would even post something like that, but it is a growing concern across many disciplines.
A problem that I can foresee is that many criminals will just not adopt this standard, and why would they? They can create their fake pictures or videos without using any kind of standard and get away with it! But I am glad to see that many major organizations are taking steps like this to provide a system/method for identifying fake images more easily.
If you are engaged in creating such content, I urge you to consider adopting such a system, to protect your content and prevent misuse!
Source: https://c2pa.org/
Leave a Reply