Your Uncle's Mac Froze After a Google Ad. Here's Why He Called His Nephew, Not ChatGPT.
The Frozen Screen Problem
Picture the scene. Your uncle clicks a Google ad, and suddenly his screen locks up with a full-bleed warning that his Mac is infected, complete with a fake support number and a countdown timer designed to make him panic. Netskope tracked this exact scareware campaign hitting 619 organizations between August 31 and September 14, running through both Mac and Windows machines, and Ars Technica broke down the mechanics on September 25.
Here's the part that matters for anyone thinking about content strategy. Your uncle didn't open ChatGPT and ask if the warning was real. He called his nephew. Or he texted a group chat. Or he posted a screenshot somewhere and waited for someone he trusts to tell him what to do.
That instinct is the whole story. When something feels urgent and wrong in real time, people reach for a human they know, not a chatbot they've never met. The scareware campaign is a small, ugly example, but it's a clean one. It shows exactly what AI search still can't replace, no matter how good the answers get. Trust in a moment of panic runs through people, not models.
Why Americans Don't Trust the Robot Yet
YouGov put actual numbers on this instinct in July. Only 28% of U.S. searchers said they trust information from an AI assistant. Traditional search engines still pull 70%. That gap is not close, and it puts America dead last among 19 markets YouGov surveyed. Not middle of the pack. Last.
The trend line makes it worse. In 2025, 82% of people said AI search felt more helpful than regular search results. By the second quarter of 2026, that number dropped to 54%. People used the tools more, then trusted them less. That is not what adoption curves are supposed to look like.
Your uncle is not an outlier here. He is the median American searcher, and the data says he is right to be skeptical. AI assistants are fast and often accurate, but accuracy is not the same thing as trust, and trust is built through repetition, accountability, and a face you recognize. A model has none of those things yet.
This is why a live human feed still does work AI has not earned. When a friend posts something in real time, corrects themselves publicly, or gets called out by a comment section, that is a verification loop happening in front of you. AI search still asks you to take its word for it. Most Americans, apparently, are not ready to.
The Citation Pipeline Nobody Talks About
Here is the part almost nobody in the "AI search is killing social media" debate wants to sit with. Social media is not losing to AI search. Social media is feeding it.
Semrush ran an analysis of 325,000 prompts across ChatGPT, Google AI Mode, and Perplexity, published in September. LinkedIn came out ahead of Wikipedia and YouTube as a cited source. Sit with that for a second. A platform built on people posting career updates and opinions about return-to-office mandates is now more citable to an AI model than the encyclopedia the entire internet was trained on.
Facebook tells a similar story. An analysis of 300 million U.S. searches found Facebook showing up in 19.5 million Google AI Overviews between July and August. That is not a rounding error. That is a platform most marketers write off as "declining" quietly becoming a source AI treats as evidence.
Zoom out and the trend confirms it. Social media's share of AI citations rose from 4.9% in January to 7.2% by August, a 47% jump in eight months. AI models are not ignoring social feeds while they mature. They are leaning on them harder every month.
So when you post something on LinkedIn or Facebook this week, you are not just talking to your followers. You are potentially handing an AI model a citation.
What This Means for Your Weekly Posting
So what do you actually do with this on a Tuesday morning when you are deciding whether to post something?
Keep posting the boring, verifiable stuff. A real review. A quick explainer about how your business handles something. A timely reaction to news your customers are already worried about, like a scareware campaign that just froze half your neighborhood's laptops. That last one matters more than it looks. When you post something accurate and specific about a real event while it's happening, you are doing the exact work both humans and AI models are short on right now.
Your uncle needed a person he trusted to confirm the scam was fake. Somewhere, a small IT shop or a solo consultant posted about that exact scareware campaign the same week it hit, in plain language, with no hedging. That post did double duty. It answered a neighbor's question and it sat there as a citable, human-verified data point for whatever crawls LinkedIn or Facebook next month.
You are not writing for followers only. You are building the raw material both channels are starved for: real people saying true things at the moment they happen. Skipping a week does not just cost you reach. It costs you a data point nobody else can create for you.