There’s an ancient symbol called the ouroboros depicting a snake consuming its own tail. It’s meant to represent infinity, renewal, death and rebirth. I believe it’s a useful image for what’s happening to information on the internet right now: We generate AI content. An LLM ingests and regurgitates it as new. And repeat.
While we can be optimistic about using LLMs for good, there’s a real concern that the cycle actively degrades quality, eroding the credibility of everything that feeds it. Research on model collapse has already shown that generative models can collapse when trained on generated data.
The giant, digital python of our internet ouroboros is decaying. But there’s an opportunity for rebirth as the snake slithers on in the form of a more accurate, more human digital space. The question is, how much time do we have left?
The Trust Problem Is Already Here
According to a January 2026 survey of 1,000 U.S. adults conducted by PANBlast and research firm Dynata, two-thirds of Americans experience AI credibility fatigue, or the cognitive drain of constantly second-guessing and verifying whether online information is accurate. Among Millennials and Gen Z, that figure rises to 76% and 80%, respectively.
This, unfortunately, aligns with Pew Research’s finding that “50% say they’re more concerned than excited about the increased use of AI in daily life, up from 37% in 2021.”’
This skepticism is all much bigger than AI-generated blog content. It’s part of an overall lack of trust across society. But change can happen in small, subtle moments that draw attention and generate engagement.
How People Are Making Decisions Right Now
When people are overwhelmed, they don’t dig deeper; they rely on shortcuts. The PANBlast research identified a behavioral pattern called ‘trust shortcuts’: quick signals people use to assess credibility without conducting thorough research.
The top shortcuts respondents reached for:
- Brand familiarity (44%): they’ve heard of this company before
- Number of reviews (35%): volume and recency of peer validation
- Google search ranking (34%): positioning within the search engine results page as an indication of authority
- Recommendations from family or friends (34%): earned word of mouth within a known circle or community
Notably, 22% said they’d use ChatGPT or a similar LLM as a trust signal. That number will likely grow, and it carries real implications for how anyone building a brand or product manages their visibility in AI-generated answers, not just traditional search. But it also raises a whole new level of concern. If an inaccurate social media claim is then “verified” with an inaccurate ChatGPT response, the person is doubly misguided.
The inverse is equally telling: 17% said survey respondents don’t trust any shortcuts at all. That’s a growing cohort so fatigued they’ve opted out of quick signals entirely and may decide that the pursuit of accuracy is not worth the gauntlet of misinformation.
Volume Is Not a Strategy
When presented with a magic button to create, the temptation to push is reasonable. Generative AI makes outright content production cheaper and faster than ever. A common consideration is to scale output. Posts, thought leadership, blogs, emails. More, more, more.
Publishing volume is only half the equation; whether you show up in the places people go to confirm a first or fifth impression is the other half.
People aren’t cross-referencing your content calendar. They’re running informal verification checks by searching your name, looking for coverage, scanning for reviews and checking whether anyone they respect has mentioned you.
Tracking referral traffic from major LLM platforms also reveals new habits: Organic citations from AI tools are increasingly functioning like referral sources. Someone asks an LLM about a category, gets a shortlist, then validates those names through traditional channels. If your name appears in that answer but lacks supporting third-party verification, the shortcut check likely fails.
How to Build Trust Without Manufacturing It
The good news is that the antidote to AI credibility fatigue isn’t necessarily complicated, but it’s not as easy as the magic “create” button.
Earn your presence in the places people already trust. A byline in a publication your audience already reads, a quote in a relevant news story, a guest spot on a podcast in your space; these function as trust shortcuts in themselves. They’re third-party editorial signals that someone independent found your perspective worth amplifying. You don’t need a massive PR operation to pursue this. You need a clear point of view and the discipline to pitch it.
Encourage your customers to do the talking. Reviews on G2, Capterra, or even LinkedIn recommendations are peer validation at scale, exactly the shortcut people are already reaching for. Making it easy for your fans to say they love you publicly is one of the highest-leverage credibility moves available. A simple, systematic ask goes a long way. I’ve found that happy clients will make the effort when asked, especially if I can convey the impact it’ll make on my role, team or across our agency.
Build a human voice alongside your brand voice. People are increasingly gravitating toward individuals over institutions. The founder who writes candidly about a hard decision they made, the engineer who shares a real technical insight, the operator who names a problem everyone in the space feels but nobody says out loud; these perspectives create credibility that no brand asset can replicate. You don’t have to be everywhere. You have to be real somewhere.
Show up consistently. One of the quieter but more powerful trust-builders is simply coherence. When someone encounters your name in a press mention, then on LinkedIn, then in an LLM-generated answer, and the story is consistent, trust compounds. Inconsistent or generic messaging fails the cross-check, even when individual placements are strong. Know what you stand for and say it the same way everywhere.
The Opportunity in the Decay
The AI ouroboros keeps spinning. The people still feeding it are competing for diminishing credibility in a channel audiences are already tuning out.
But the decay creates an opening. In a sea of undifferentiated, algorithmically optimized content, specificity is a differentiator. Genuine expertise is a differentiator. A real human perspective, even an imperfect one, is a differentiator.
I don’t believe the internet is broken beyond repair. It’s being sorted. The sources, voices, and organizations that built real credibility distributed across earned channels, peer validation, and consistent human presence will emerge as the trusted shortcuts of the next cycle.
Cited links
- Pew Research Center: Americans remain more concerned than excited about AI in daily life. https://www.pewresearch.org/science/2025/09/17/views-of-ais-impact-on-society-and-human-abilities
- Nature: Research on model collapse from recursively generated data.https://www.nature.com/articles/s41586-024-07566-y
- PANBlast/Dynata survey on AI credibility fatigue and trust shortcuts. https://www.panblastpr.com/resources/trust-issues-ai-credibility-ebook/
