A striking new study from researchers at Stanford University, Imperial College London, and the Internet Archive reveals just how fast artificial intelligence has woven itself into the fabric of the internet. According to the analysis, first reported by 404 Media, 35% of all web pages created between August 2022 and May 2025 contain AI-generated or AI-assisted content. That’s one in three new pages.
The shift began almost overnight. The researchers tracked archived pages over 33 months and found a sharp inflection point right after OpenAI’s ChatGPT launched in November 2022. By mid-2025, the share of AI-touched pages had climbed from near-zero to roughly 35% – and the curve is still pointing up.
Beyond the headline number: How AI is reshaping online tone and trust
The study didn’t just count pages. It measured language patterns, and the findings are just as consequential for leaders and communicators.
1. The web is getting blander – and more positive
AI-assisted text tends toward upbeat, straightforward phrasing with less stylistic variation. For brands, that means a rising “sea of sameness.” The risk for Nordic companies is that differentiated voice – a key asset in markets that value authenticity – gets diluted. In B2B, where trust is currency, sounding like everyone else is a strategic liability.
2. Not more lies, but fewer facts you can check
The researchers looked for an uptick in verifiably false claims and didn’t find one. But that’s not the whole story. As co-author Jonáš Doležal of Stanford told 404 Media: “It could still be that AI is quietly increasing the number of unverifiable claims – those that cannot be checked against existing fact-checking tools and infrastructure.” Or, as Doležal put it more bluntly: “It could simply be that the internet wasn’t a place that stuck to the truth very well to begin with.”
For decision-makers, the takeaway is clear: The problem is shifting from outright misinformation to a flood of plausible, citation-light content. Due diligence just got harder.
Analysis: Why this matters for Nordic business in 2026
The cost of credibility is rising
In Sweden, Norway, Denmark and Finland, public trust in institutions remains relatively high by global standards. But if 1 in 3 web pages is AI-assisted, buyers and partners will default to scepticism. Expect procurement teams to demand more original research, proprietary data, and human-authored thought leadership. “AI-washed” content will convert less.
SEO and visibility rules are being rewritten
Google’s March 2026 core update explicitly targets “scaled, low-value AI content.” Sites that publish high volumes of generic AI text have seen 20-40% traffic drops in Q1 2026 Google. Nordic exporters relying on inbound marketing should audit their libraries now. The winners will be those blending AI efficiency with subject-matter depth and local nuance.
Labour market implications
The study period ends May 2025, but LinkedIn data from Q1 2026 shows “AI content editing” is now the fastest-growing skill in the Nordics, up 78% YoY. The job isn’t disappearing – it’s shifting from writing to verifying, sourcing, and adding judgment AI can’t fake.
Regulatory angle
The EU AI Act’s transparency requirements for “synthetic content” began phased enforcement in February 2026. While the law doesn’t ban AI web copy, it requires clear labelling in some contexts, especially ads and political content. For Nordic companies operating cross-border, compliance will soon be a web governance issue, not just a legal one.
3 moves to make this quarter
| Action | Why it matters now | How to start |
| Run an AI-content audit | Investors and customers will ask how much of your site is machine-written | Use detection tools + manual review on top 50 landing pages |
| Double down on provenance | Unverifiable claims erode trust faster than lies | Add named authors, dates, primary sources, and original data to key assets |
| Train for “AI + Human” workflows | Bland content doesn’t sell in high-trust Nordic markets | Upskill teams on editing, fact-checking, and adding field expertise to AI drafts |
The bottom line
AI didn’t break the internet’s relationship with truth – it amplified a trend already underway. But it did accelerate the commoditisation of content. In the next 18 months, the premium will be on what AI can’t generate: experience, accountability, and verifiable insight.
Nordic firms have an edge here. Our business culture already prizes transparency and understated expertise. The challenge is to make that visible online before the web’s rising “positivity tide” washes it out.
Next in Nordic Business Journal:
In our May issue, we’ll dive deeper into “The New Rules of SEO in an AI-Saturated Web” – featuring interviews with CMOs at three Nordic scale-ups that grew organic traffic 60%+ after the March 2026 Google update. We’ll break down their playbooks and share a template for your own AI-content policy.
Connect with us:
Have you audited your website for AI-generated content? How is your team balancing efficiency with credibility? Email the editors at insights@nordicbusinessjournal.com or join the discussion on LinkedIn and Instagram. We read every message and may feature your perspective in our next issue.