Our previous coverage documented AI hallucinations polluting published books and award-winning fiction. Now the damage is moving upstream — into how people process information itself.

A 2025 analysis from SBS Swiss Business School found a significant negative correlation between frequent AI tool use and critical thinking ability — across all age groups and education levels. The more someone relies on AI, the worse they test on independent reasoning.

Younger users fared worst. The research showed Gen Z exhibits the highest AI dependence and the lowest critical thinking scores of any cohort studied.

A Pangram/YouGov study published in May documented the detection problem: only 55 percent of Gen Zers aged 18 to 28 could identify fake or misleading AI-generated content. Older adults did worse. A majority of American adults over 28 are not confident they can spot AI-generated misinformation.

Volume Is Being Mistaken for Consensus

Javi Pérez, an editor of AI-assisted consumer education websites, told The Epoch Times the mechanism at work: volume mistaken for credibility.

"If a user sees dozens of similar posts about a product, trend, political claim, health issue, or financial topic, they may assume there is broad agreement," Pérez said.

He calls it "confident sameness" — the way AI-generated content repeats similar structures, similar advice, and similar phrasing across hundreds of articles until readers assume consensus exists where it doesn't.

In health, finance, law, and political coverage, this carries real consequences. Readers have no way to know if a claim was reviewed against primary sources, updated recently, or edited by anyone accountable.

Stanford Already Flagged This — Nobody Acted

A 2024 report from Stanford's Human-Centered AI Institute found that more than 30% of online text already shows signs of AI assistance, according to AI News International. That study came out over a year ago.

The flood has accelerated.

AI strategy consultant Armand Cucciniello III, who described his background in the U.S. national security landscape to The Epoch Times, observed: "We're moving from deliberate reading toward rapid skimming of polished summaries, commentary, short-form videos, and AI-assisted content designed for speed and engagement."

The platforms are designed to reward exactly that behavior.

What Mainstream Tech Media Is Getting Wrong

Most tech coverage frames this as a content quality problem — a nuisance to be solved by better AI detectors or Google algorithm updates.

The SBS Swiss Business School research isn't saying AI is producing bad articles. It's saying AI dependency is restructuring how human brains engage with information. Detection tools don't fix that. Better prompts don't fix that. This is a behavioral and cognitive shift happening at population scale.

Google's "helpful content" guidelines, cited by AI News International, have limitations. Google has financial incentives to keep AI-generated content flowing through its ad ecosystem.

The National Security Angle

Cucciniello's national security background is relevant for a reason. Adversarial actors — China, Russia, Iran — don't need to plant specific lies when the entire information environment is already saturated with AI-generated "confident sameness." Confusion and distrust become the weapon. An electorate that can't distinguish real from fake is an electorate that's compromised.

China understands this dynamic. Washington's response remains unclear.

What This Means for You

If you're a parent, your kid is growing up in an information environment where the majority of their peers can't reliably tell what's real. Schools aren't teaching media literacy. Platforms aren't fixing it. Government isn't acting on it.

If you're a voter, the political "consensus" you're seeing online may be manufactured at scale by systems with no accountability.

If you're a consumer making financial or health decisions based on what you read online, that information pipeline is now majority-suspect.

The old rule was "consider the source." The new problem is there are no sources anymore — just an undifferentiated flood of polished, confident, machine-generated text that all sounds the same.