North Carolina Lt. Governor Mark Robinson, the current Republican candidate for governor, has vehemently denied claims made in a now-famous CNN report – he called them “salacious tabloid lies” – and Robinson accused his opponents of using AI and other digital tricks to “manufacture” false evidence against him.
He argued that these comments – such as saying he is a Black Nazi who enjoyed transgender porn – didn’t reflect his character and said in an interview that one of his adversaries had spent a million dollars to discredit him using the help of AI, Artificial Intelligence.
An expert computer programmer who’s worked for major companies for a quarter of a century told the Rhino Times this week that to alter past posts on websites that one doesn’t own or control, one would need access to the websites either with the help of high-level administrators or an extraordinary hacking effort that could break the best security systems in the industry, such as those used by Amazon or Twitter, where some of the incendiary posts – allegedly made by Robinson – occurred.
While the Nude Africa website Robinson allegedly posted to might not have top-notch security, some of the websites on which Robinson allegedly posted under the alias “minisoldr” are Amazon and X, which at the time, about 15 years ago, was known as Twitter.
The programmer noted that while undertaking such a massive effort spanning many sites might theoretically be possible, he questioned why anyone would put the incredible amount of time and resources into a takedown effort like that against a candidate who looked as though he was going to lose badly in the election anyway.
Robinson was down 11 points before the CNN report and, in the wake of it, he is trailing 17 points behind his Democratic rival Josh Stein.
“Why wouldn’t they go after a higher, more compelling target,” he said, “if someone were able to do something on this scale?”
The programmer also said that due to constant security threats to major sites, the type of top computer administrative authority, to say, change time stamps on posts, is only given to a select few. He said for instance that he had been working for his current employer for two years and was highly trusted but still does not have the credentials or the ability to alter his company’s website in a way that Robinson is claiming happened.
Chat GPT, the highly popular artificial intelligence engine, also responded that it would be an extremely, extremely, hard thing to fake.
The AI engine, when asked, stated that, “While the claims rely on matching personal details and an email, Robinson’s defense hinges on the notion that the evidence could be fabricated digitally. In theory, falsifying online identities and comments isn’t impossible, especially with AI tools, but it would be challenging to fake multiple detailed biographical links that span years, which CNN claims to have matched.”
“Given this,” Chat GPT added, “proving whether or not the evidence is entirely falsified would require a thorough forensic analysis of the digital footprints involved.”
When Chat GPT was asked why it would be so challenging to fake multiple detailed biographical links that span years, here was the Artificial Intelligence engine’s response:
Faking multiple detailed biographical links over several years would be challenging due to the complexity and consistency required in such a fabrication.
Here are the main reasons:
Digital Forensics: Online interactions leave digital footprints that are traceable across different platforms, servers, and IP addresses. It’s difficult to manipulate these traces without discrepancies. If a user’s identity (like an email address or username) is repeatedly connected to specific posts over several years, a forensic examination could track and confirm the authenticity of those connections. This would include time stamps, geographical data, and platform usage patterns.
Long-Term Consistency: Fabricating a set of posts across years requires maintaining a high level of consistency in writing style, tone, and specific personal details. If someone were trying to fake this, they would need to ensure that the posts align with known aspects of Robinson’s biography at different points in time, making it more complex than a one-off forgery.
Multiple Data Points: CNN reportedly linked Robinson to the comments using multiple biographical details and an associated email. Forgeries involving multiple layers of personal data—like matching personal events, language use, and specific life details over a long span—are harder to create without being flagged for inconsistency.
Historical Records: Websites and forums often store historical data that third parties can access. Creating a false history that goes unnoticed by these platforms, or that withstands the scrutiny of years’ worth of independent verification, would require considerable effort and precision…In Robinson’s case, CNN claims to have matched not only posts but also other personal identifiers, like his email, to create a consistent profile over years. While theoretically possible to fake such evidence with modern technology (like AI), doing so accurately and in a way that avoids detection would require immense resources and expertise.”
The good news for Robinson is that, if the CNN report is false, Robinson can launch a lawsuit against the network that will make him a multi-millionaire many times over.
That’s rich. A man’s life could be ruined, yet we are allowing one AI tell us that another AI wasn’t used, to create posts. Scary times.
I read Chat GPT also said there was no credible evidence someone tried to assassinate Oresident Trump, so there’s that.
When I studied computers there was a catch phrase- ” Garbage in, garbage out….”
Obviously, artificial intelligence is only as intelligent as it’s told to be.
ChatGPT’s information source is the internet. So tons of garbage for sure.
Thanks so much, Scott Yost, for sharing just how difficult it is to fake a third party posting on the internet. Your article is very sobering. Thanks for including AI’s opinion, as well, considering AI was inculcated in the story.
We understand the appeal Mark Robinson has on the Greensboro community. He had tried to show a path for anyone with a political opinion and a voice that it was possible make it into politics and make a difference.
Thank you Scott for this article, it proves that Mark is human and makes mistakes. My dad taught me that if a man makes a mistake and then apologizes…you accept that apology.
Mark made a mistake and LIED about it.
This guy is unfit to be governor
Excellent investigative journalism. You may lose some readers of the MAGA persuasion who refuse to consider facts that may not support their steadfast beliefs.
Well I’m a MAGA conservative (with a libertarian bent) and I considered the allegations credible and persuasive from the beginning. I said as much in these columns.
Perhaps your liberal prejudices are showing, Mr Bretzmann…
Win!!
“one would need access to the websites either with the help of high-level administrators or an extraordinary hacking effort that could break the best security systems in the industry, such as those used by Amazon or Twitter” (I wonder who that could be?)
Given the last few years, I don’t know what to believe or trust. I know that releasing this info on Robinson the day before NC closed options for Robinson’s name to be removed from the ballot and another GOP candidate to be substituted is suspect.
I suspect the real target was Trump. They knew fewer votes for Robinson meant fewer for Trump, and NC is a critical state for Democratic votes. If they could put out dirt on Robinson at the last minute, with no time to prove it was false, they could make enough people hesitant to vote a straight GOP ticket, which most do. We’ll see.
You are so right, Deborah.
This has the fingerprints of the DNC all over it.
Lol, I think this has the fingerprints of Mark all over it, and his computer, and his…. LMAO
You like laughing at your own jokes, don’t you?
You should be a comedian.
Right! The implication and timing suggests that the DNC is behind the whole thing. Basically to drag down Trump.
This article is all predicated on the assumption that:
A. All the posts are fake. When in fact, some of them, like on the legitimate sites could be his. Then someone used the publicly visible user name to post on the illicit sites and mimicked the publicly available biographic info and writing style. Not too hard to do.
B. That Nude Africa, which I assume isn’t the most moral company, is not directly in on it. They could have simply been paid a sum of money to allow access to their site’s inner workings. No hacking needed at all, just a simple bribe. Imagine all the extra hits their site is getting from the press they’ve received. Certainly nothing for them to gain here…..right.
C. That any AI system is any more reliable than Wikipedia. AI has already proven to be unreliable. It is as biased as it’s programmers and only marginally reliable when asked to make interpretations. Half the time when you ask it something, it’s isn’t even replying using current data. For example, I asked it who would win today’s football games and it gave me predictions based off of the 2023 schedule. Not reliable at all.
Do better. I’m not saying the dude is innocent, but this “proof” from AI is a joke.
Why would someone fake hisbuser name from ten years ago and link to her personal life? Lol he was a nobody ten years ago.
And, no, you can’t just ‘fake’ comments to look tens years ago. Keep in you would have to fake the entire range of other profiles he was responding to and those responding to him
That would require a massive effotlrt that would basically require the entire site to be faked and manually created to bring down a guy that already lost given recent polling.
Keep bending over backwards to protect this guy and soon your head will be up your ….
But you be you.
Again, you have it wrong. During every political season ads using AI are changed rapidly to put that candidate in a bad light and it has been happening since the early 2000s. And mainly for down-ballot candidates. Remember. we are talking AI. All it takes is good computer programmer to create a program which will filter words from a candidate then create an avatar for it to see those words completely out of context. This is done weekly by those who create parody ads and jokes. Stop showing your political leaning by printing such rubbish.
AI cann’t manipulate a web servers data stores going back a decade. That’s not what AI does.
Mark is blaming AI because it’s a buzz word and I highly doubt he knows much about technology given his open use of these websites given the lack of privacy of anything you do online.
If it is possible to even find WRITTEN posts on an obscure porn site from 2008, how about someone do a deep dive into Josh Steins Internet history during that time? He associates with some pretty strange and demented characters, so I expect it to be pretty disgusting.
Allegations… Schmelegations.. He is MAGA thru and thru so he gets a pass. He’s Trump’s man. We vote for whoever says Trump says. It’s 2024.
Vote for Mark Robinson and Donald Trump.
if you don’t care about truth or facts in your support of conspiracy theories, and lack morals and sense of decency, Vote Trump and Robinson. You will fit right in.
Otherwise, Go Blue!
this is proof of a ‘deep state’ that spys & archives our every move & ever-ready to blackmail , bully or shame when advantageous to them. as predicted by g orwell.
Too late for Republicans to control the damage,Trump failed to vet this guy before he endorsed him. Trump teaches the maga crowd who to hate. Robinson continues the tired hate message.Looks like these two are kinda stuck together.How ironic that this guy could help sink Trump last effort.
P.S
somebody show this guy how to get out of a truck without burning his finger prints off
Oh well, everything is alright, then.
this is proof of a ‘deep state’ that spys & archives our every move & ever-ready to blackmail , bully or shame when advantageous to them. as predicted by g orwell.