The power of artificial intelligence is amazing. I recently found myself bored and cruising the news feeds. I ran across a disturbing story of alleged criminal activity involving two minors. The habit in this country has long been to identify youth figures only with their initials.
This was an evolution of the 20th century, which I recently learned some young(er) folks now refer to as the "1900s," even when they are discussing as late as 1999. The halcyon days of my youth are thus "so last century," and that both educates and depresses me.
The specifics in this instance involved two young people and made the news due to a civil lawsuit filed by the victim. The story said the assailant and victim had each transferred to different schools since the event, that there had been an arrest and prosecution, and other specific details. One critical detail named a parent/guardian with no similar effort at obfuscation identity. Details of the victim's school activities participation, and some physical description were included.
I was intrigued immediately because of a presentation at the WCI on artificial intelligence. That expert had explained that AI is much stronger and powerful than the algorithm searches (internet searches) to which we have become accustomed.
Thus, the internet search allows us to look for keywords. I used that to search for the initials mentioned, the town referenced, and other details. I found news stories and similar references. They were largely homogenized, consistent, and even redundant-A.B.C alleged to have been attacked by X.Y.Z. It struck me that I might identify either the victim or the perpetrator.
The articles described activities (sports) in which each was involved, as that formed some foundation for the described actions and allegations. I therefore took a chance on a reasonably simple AI prompt, not about X.Y.Z., but about the activities. Local news loves to publish names and even pictures of young people excelling at academic and athletic events.
I asked, "Compile me a list of all students in _________ participating in ____________ in MONTH of YEAR." I received a disappointingly brief response with an offer to dig deeper. I asked for the "dig deeper," and after several such offers and responses, I was provided an extensive list of students who were known for participation in those activities in that community.
Only one name had any potential to match X.Y.Z.'s initials. In fairness, this was aided by the fact that the first initial (no, it was not really "X") was not radically uncommon, but was also not at the top of the common initial list (the predominance of the top 9 is amazing—J, M, A, C, D, R, S, L, and B account for the vast majority).
A follow-up prompt quickly confirmed that this named student now attended a school some miles away, which matched the generalities in the original news article. The path to photographs, background, and more was then reasonably unchallenging. Was all this enough to convince or convict? Unlikely. Nonetheless, the inquiry is reasonably supportive of the alleged perpetrator's identity.
A similar prompt led me reasonably rapidly to a probable identity of the victim. Armed with the information, the power of AI was reinforced. The ease of pulling back the protections of "initials" was frankly unsettling. The fact that it was easy for a dinosaur like me was more so.
I have often related the story about the pace of regulatory change. I first heard it in 2015, as discussed in Salim Ismail and a Life-Changing Seminar in Orlando (May 2015). The speaker noted that most states have laws requiring rearview mirrors on all automobiles, but none require a steering wheel. The simple fact is that regulators did not think of that when drafting laws. Who would build a car without one? Well, welcome to the 21st century (or "it ain't the 1900's anymore").
So, the point of the foregoing is not to shame either victim, perpetrator, or others. I am doubtful that there is enough shared here for the reader to even guess at the news story that brought me here (there is a challenge for your AI skills). Instead, the point is that the time has likely come for a more effective method of pseudonym for the protection of minors in legal proceedings. For centuries, courts have striven in this regard.
See the history in Lupton, Illinois Supreme Court history: John Doe and Richard Roe (). If the effort is sincere, the path merits attention. Identity is increasingly revealed or open in the age of the internet. The advent of artificial intelligence will only make it more so.
Beyond the pseudonym, there would have to be some method to diminish access to various details, such as parent names, specific school or community activities, and more. One reaction is that this is an affront to the free press, another that this would be significant work, and there are more arguments to make.
Nonetheless, if it is worth the effort to protect the identity of minors, it is worth a better effort than this experience demonstrated. If it is not worth the effort, then the superficial A.B.C. and X.Y.Z. are now likely of little or even no use whatever. The decision seems to be whether it is worth doing at all. If it is, is it worth doing effectively? I suspect that A.B.C. and X.Y.Z. both think it is.
