Image Source: Nano Banana Pro - Google Gemini
Outline
- Intro: Question Reality
- Network Chuck video - please take 20-minutes to watch this!
- Detection Tools and Tips
Question Reality
We have been inserted into a battle we cannot win, at least, without awareness and tools.
While AI has been around in various aspects for decades, Generative AI went mainstream in late 2022 when OpenAI released ChatGPT. At its initial appearance, it was quaint, quirky, and qualitatively obvious. We could have it write something or even create crude images that were painfully obvious that they were AI generated.
In three short years, it has improved so drastically, it is impossible in many cases, to detect AI generated content from human created content. At the moment, the #1 song on the Country Top 40, is by AI.
A small example of an AI generated image that appears convincingly real, is the banner image for this post. Nano Banana, a tool by Google Gemini updated a couple of weeks ago and is mind-bending powerful.
The image was generated in a few seconds using a prompt:
It is unsettling to me that this technology has advanced so quickly, even reading and studying it constantly, it is virtually impossible for me to keep up. My desire is to engage, educate, and empower others to understand this technology and be able to protect themselves. See video below.Create a hyper-realistic, ultra-sharp, full-color large-format image featuring a massive group of celebrities from different eras, all standing together in a single wide cinematic frame. The image must look like a perfectly photographed editorial cover with impeccable lighting, lifelike skin texture, micro-details of hair, pores, reflections, and fabric fibers.GENERAL STYLE & MOOD: Photorealistic, 8k, shallow depth of field, soft natural fill light + strong golden rim light. High dynamic range, calibrated color grading. Skin tones perfectly accurate. Crisp fabric detail with individual threads visible. Balanced composition, slightly wide-angle lens (35mm), center-weighted. All celebrities interacting naturally, smiling, posing, or conversing. Minimal background noise, but with enough world-building to feel real.THE ENVIRONMENT: A luxurious open-air rooftop terrace at sunset overlooking a modern city skyline. Elements include: Warm golden light wrapping around silhouettes. Polished marble.
Network Chuck
Summary:
My Highlights:
If you talk to law enforcement, which also has access to the actual databases of people who paid, the real number is closer to 70%. Think about that. Small businesses, families, grandmothers—your grandmother—paying $600 or more that they cannot afford.
Voice-cloning scams:
We have seen voice cloning used in scams, especially ones attempting to mimic family members or friends. It only takes eight seconds of your voice to clone it. Imagine your mom getting a call that sounds exactly like you saying, “I’m in trouble. I need money.”
AI honeypots catching scammers:
In the video, Roz calls an AI honeypot pretending to be a scammer. The AI plays the victim—and even has personality. It proceeds to engage the scammer just like a real person would.
Multi–touch-point attack detection:
Scammers now start a scam on WhatsApp, then continue it in your browser, then switch to phone. Banks added obfuscation so screens shared from a phone show as blank, but scammers adapted: “Don’t do it on your phone. Open it in your browser instead.”
Heartbreaking ransomware case:
A couple in Syria had their single family computer infected. They could not pay the $350 ransom—which may as well have been $35,000. What made this case devastating was that the attackers were holding hostage the only remaining photos of their two children who had died in the war. These criminals were extorting them using the last memories of their children.
Cybersecurity as a human right:
Cybersecurity should be a fundamental human right—just like liberty and critical thinking. It should not be conditioned by money. Once people realize how deeply digital life affects their physical life, they understand the need to secure it.
Detection Tools and Tips
- Be careful what you post on platforms without required authorization - we will talk about OSINT (Open Source Intelligence) in future posts
- Use pass phrases instead of passwords, e.g. "cow jump moon" (don't use that) is much more secure than "Winter2025!"
- Use multi-factor authentication everywhere (NEVER give your code to anyone)
- Use a password manager, there are a lot of great options, a couple of my favorite, easy-to-use options are:
- BitWarden - free options, cross platform
- Apple Password manager - built in to iOS devices
- Use passkeys where available
- These use biometric authentication (fingerprint, facial recognition, etc.) reduces friction and improves security - win win!
- Question everything
- Use extreme caution when clicking on anything, opening attachments - double and triple check
- Always verify - never call a number in an email, look it up in Google or other source and verify legitimacy
- Have a family code word - yes, it is crazy we are here, but in the event of a deep fake voice attack a code word will be helpful
- NEVER give an unknown person remote access to your computer
- Google, Microsoft, the IRS or anyone else will call you out-of-the-blue and request access to your computer
- If they do, hang up. Look up their number elsewhere and call them back
- Deep Fake verification tools - these are new and not completely dependable, but a good first step
- https://scanner.deepware.ai/
- https://deepfakedetection.io/deepfake-image-detection
- Make backups of your important information and keep it disconnected from your network; use cloud storage
- Incredibly good, easy-to-use tool to work through improving your home security https://digital-defense.io/