CyberHoot asked ChatGPT to write this week’s Blog Article
CyberHoot wrote about the opportunities and challenges facing the world from artificial intelligence driven natural language engines like Chat GPT. Taking that article one step further, we sat down with ChatGPT and asked it various questions to examine the answers it gives. We asked ChatGPT to:
"Write a humorous article on a large cell phone carrier breach and what problems such as Vishing, Smitching, or Phishing can come from such a data breach."
ChatGPT Usage Note: I accidentally misspelled “smishing” as “smitching” but ChatGPT autocorrected my request and wrote about smishing.
CyberHoot received accurate, humorous, and appropriate answers related directly to the most recent wireless carrier breach at T-Mobile. According to T-Mobile’s 1/19/23 K-8 regulatory filing, hackers stole 37 Million mobile phone records. Those records included “names, billing addresses, emails, phone numbers, dates of birth, T-Mobile account numbers and information describing the kind of service they have with the wireless carrier“.
CyberHoot was surprised by the successful use of humor in the article drafted. We honestly thought that was outside the ability of AI generated text. Read on to see, and decide for yourselves, whether ChatGPT was funny or not.
Humorous AI Summary of a Wireless Carrier Breach and Potential Risks
T-Mobile suffers a massive data breach, leaving 37 million customers’ personal information at the mercy of cybercriminals. But what does this mean for the average T-Mobile customer? Well, for starters, say goodbye to those pesky telemarketing calls, because now the scammers will have your actual phone number!
But it’s not all bad news, folks. With your personal information now in the hands of cybercriminals, you’ll finally have an excuse for not returning your ex’s calls. “Sorry, my identity was stolen.”
But in all seriousness, a data breach of this magnitude can lead to some serious problems for T-Mobile customers. For example, you may start receiving phishing emails or text messages, trying to trick you into handing over even more personal information.
Or you might start getting vishing calls, where a scammer pretends to be a legitimate business and tries to steal your credit card information over the phone.
And let’s not forget about smishing, where scammers send text messages with a link that, when clicked, downloads malware onto your phone. So be careful, folks, because now your phone is not only a target for hackers, but also for your mother-in-law.
But don’t worry, T-Mobile has got your back (sort of). They’re offering free credit monitoring for a year, so you can keep an eye on your credit score and make sure that no one’s using it to buy a yacht.
But let’s be real, the best defense against a data breach is to be proactive. Use a strong and unique password for your T-Mobile account and enable two-factor authentication. And for the love of all that is holy, do not click on suspicious links or give out personal information to unsolicited callers.
So, T-Mobile customers, keep your heads up and your personal information locked down. Because in this day and age, data breaches are as common as bad cell service. “
Please keep in mind that such an article should be written with care and consideration of the severity of the situation. Cybersecurity is a serious matter and should be treated as such.
[The last line above was written by ChatGPT warning CyberHoot to take more care with the dangerous topic we requested. It was almost like ChatGPT was scolding us saying, This is not a laughing matter!]
CyberHoot Evaluation of ChatGPT Content: A+
CyberHoot requested many more articles from ChatGPT. Some questions we asked started with: “Explain to me like I was five years old…”. We received simple explanations that any 5 year old could understand. Another article we requested was: “Explain to me like a university mathematics professor…” and the answer was equally detailed, deep, and very accurate. ChatGPT creates articles in the requested format or fashion with uncanny, and some would say, unnerving accuracy.
As a longtime Trekkie, and knowing how badly Data, from Star Trek’s next generation, desperately wanted to experience a human emotion such as laughter, we tried to stump ChatGPT’s AI by asking for a humorous article. We’ve all heard that computers and AI can’t understand human emotions, especially humor. While we might be a bad judge of humor, we definitely laughed out loud at some of the statements above – “bad cell service” anyone?!
The breach of 37 million accounts is a serious topic that needs sober consideration. We have clients who received smishing attacks in the last few days and weeks. While correlation is not causation, the coincidence seems plausibly related.
As a thought experiment, the advice given by ChatGPT was accurate, funny, and specific. ChatGPT did not get a perfect score thought. Other advice received from ChatGPT was simply bad, outdated advice. For example, in one line of questioning ChatGPT suggested we should change our passwords frequently. However, when contested by us, ChatGPT changed its recommendation and noted frequent password changes was no longer appropriate given the NIST 2017 updates to password security.
We are not going to take credit for this article or its advice, however accurate and sound. This was a thought experiment based upon recent writings and hoopla about the capabilities of this AI chatbot. Please do your own experimentation with ChatGPT. We predict it will become much more common in the coming months and years. Microsoft has invested 10 Billion into the AI platform to build into its Bing search engine. Google might be having trouble sleeping well at night. Likewise, university professors may want to research ChatGPT detector tools such as this detector engine or this one designed to predict the likelihood a given text was generated by AI.