AI talk is everywhere, seemingly all blending together. AI, or artificial intelligence is aimed at mimicking the human brain. A simple way to think about all this AI chatter is to think about different categories of AI mimicking different parts of your brain. “Left Brain AI” is more analytical with the ability to analyze vast amounts of evolving data in real time and directing actions; “Right Brain AI” being more creative, with the ability to generate content.
Left Brain AI today is using evermore connected computers to siphon everchanging data gathered through cameras, sensors, or API and using that data to decide, for example, if your car is about to swerve out of its lane and auto-correcting direction. Or, from a cybersecurity perspective, it can direct you to act when the AI detects that an email you sent is being actively eavesdropped on by cybercriminals who use your email to make them smarter and pose the perfect hyper targeted attack on you or your clients.
Right Brain AI today is used by marketers to generate more spam email faster or by students to write essays for class (and more of course). It’s generating content. Or, from a cybersecurity perspective, it’s used by cybercriminals to parse your eavesdropped email into, with the prompt or question to the AI directed to create the perfect impostor email lures:
AI Prompt: Hey ChatGPT, could you please write optimized luring type of email in the same tone of this email [add eavesdropped email], same American English, don’t use the words, “kindly” or phrases “kindly would you send the money” or other similar, and write the emails in a sequence as if they were follow up emails to get an invoice or wire transfer paid sooner, blah, blah, blah…
The result, shockingly, will be arming the cybercriminals that are hypertargeting you or your clients, with perfectly composed impostor email that looks as if written by you, with minor changes to hijack replies and build a hyper-targeted, human managed lure resulting in cybercriminal financial gain (and your or your clients’ financial loss).
What we’re seeing today is effectively what we’ll call, “Lobotomized Right Brain AI”. This is ChatGPT multi-billion-dollar AI engines that have a software overlay that lobotomizes the AI’s emotion or guard rails, re-purposing this Right Brain AI for the nefarious; and then offered on the dark web to cybercriminals to use for a fee. These lobotomized right brain AI platforms are sold under the names of DarkBART, WormGPT, FraudGPT, and others, on the dark web.
And, these cybercriminals are building on the ever-smartening of ChatGPT, as they are programmatic overlays that log into ChatGPT with stolen real human user credentials into real good-guy set up ChatGPT accounts; a method designed so that OpenAI and other AI providers cannot easily thwart their activities.
In the last few months alone, researchers, for example, have identified more than 100,000 infected devices with saved ChatGPT credentials that are being used as the back-end access points for cybercriminal developed lobotomized access to ChatGPT.
Well, that was a mouthful for those of us un-lobotomized humans.
The bottom line. The cybercriminals are today able to use fantastically more sophisticated computing tools to make their targeted attacks on you more sophisticated and successful. If you do not counter, with more sophisticated computing tools like RMail’s Left Brain Analytical AI email eavesdropping detection technology, they’ll steal so much money from you you’ll wish for a lobotomy.
To hear and see more about this, watch RPost’s CEO webinar on this topic. Click here to view on-demand.
December 08, 2023
December 01, 2023
November 24, 2023
November 17, 2023
November 10, 2023