I’m vigilant in this current internet era where scams are plentiful and everyone is a target. From 2024 to 2025, the Internet Crime Complaint Center reported a 17% increase in filed claims. I’d be willing to bet that we’ll see even more complaints by the end of 2026, especially with the rise of AI and its increasing accessibility. Text message scams have overtaken email, and scams involving voice and video cloning have cost Americans almost a billion dollars. The actual number, accounting for unreported cases, is likely much higher.
AI Scams Are Evolving
1️⃣ Fake Creators Pushing Crypto Coins
An account that has blown up over the last few weeks is Jean Phil. Shortly after the account gained traction, it started promoting a meme coin. The profile has since removed the link (and deleted the original page), but it doesn’t change the fact that an AI influencer quickly gained a massive following and started promoting a cryptocurrency.
2️⃣ AI Testimonials and Products
One of the ways AI scams have been showing up on my feed is through deepfaked product reviews. It initially started with copying the likenesses of celebrities (you may remember when people were creating deepfakes of Joe Rogan endorsing products), but now I’m actually seeing a trend where AI-generated people who were never real in the first place are making fake testimonials for brands and companies.
3️⃣ Voice Cloning Scams
Maybe the most nefarious ones on this list, these scams involve AI cloning real people’s voices and using them to extort money. Last week, scammers cloned the voice of a sheriff in Osceola, FL, claiming that residents had missed court dates and needed to pay a fee or risk going to prison. This is an extreme case of newer AI technology being used to exploit old scams, but I don’t see this trend going away anytime soon.
Should Platforms Be Doing More to Police AI-Generated Content?
I’m pretty skeptical that this is going to happen, because if you remember when AI content started to ramp up on Instagram, they had a prompt before uploading content asking if the person (or bot) uploading wanted to add an AI tag to the post. Anyone could just skip this step. Then Instagram started adding “AI Info” badges, or AI content tags, but this seems completely arbitrary because none of Jean Phil’s videos include any kind of AI content tag.
Content marked AI
My other concern is that if platforms start flagging more AI content, it’s inevitable that their systems are going to incorrectly tag content that wasn’t made with AI as AI-generated (which has happened in the past). Even if every piece of AI content was correctly marked with an AI tag, I still think people would glance over it. I don’t think social platforms are in any rush to fix any of this, which will only create more confusion when doomscrolling at 3 in the morning.
The Big Takeaway
AI is making it easier to create fake people, promote crypto scams, and impersonate real people to steal money. As these scams become more sophisticated, social platforms are struggling to keep up. AI labels are easy to bypass, inconsistently applied, and can even flag real content by mistake.
If you enjoy reading Night Light each week, refer a friend and get some Night swag, including a personal video from me, an invite to our Discord, and I’ll even put your picture up on the walls of our office forever.
Everyone is trying to get their hands on the new iPhone Duo. Shoutout to JerryRigEverything for completely destroying his by pouring sand all over it. Good this this gif doesn’t have sound, because it’s crunchy.
Hollywood Reporter Drops Top Managers List
Congrats to everyone on the list.
Elizabeth Holmes Lookalike Contest in NYC
You Can See Everything has a 3 hour runtime, not long enough.
