Seven Myths About AI Search, Disproved by Data
Analysts have tested popular claims about AI search. Traditional search is growing alongside AI, and 97% of visits following mentions in chatbots go untracked in analytics.
AI is not killing traditional search
Open LinkedIn on any day and you'll find someone claiming that traditional search is dead, that AI content no longer ranks, and that SEO teams are obsolete. Most of these claims sound plausible and get repeated until they turn into "common knowledge." But check them against data, and they fall apart.
Myth 1: AI is killing traditional search. The claim goes: every session in ChatGPT, Claude, or Perplexity is a search that Google didn't get. The State of Search Q2 2026 report (Datos) showed something different: AI search and traditional search are growing at roughly the same pace quarter over quarter. AI platforms aren't cannibalizing search demand — they're building on top of it. The nuance is that clicks to websites are indeed falling: Google increasingly answers the query without sending the user anywhere. The share of clicks on external results on desktop dropped to its lowest point since April 2025. Demand is growing, but Google is sending fewer people to the open web.
Organic traffic grows even in the zero-click era
Myth 2: in a zero-click environment, organic traffic cannot grow. Fact: one of the author's clients recorded a record month for organic visits in July 2026 — entirely through non-branded efforts on blog and service pages. It's a local business in a different risk category, and not every site will replicate that result. But the accurate framing is this: zero-click pressure is real, but it's unevenly distributed across industries, and a smart SEO strategy still delivers wins.
Myth 3: fully AI-generated content doesn't rank. The author's test disproves this: four articles, fully generated by AI with careful prompts and published from March through September last year, continue to rank and drive traffic to this day. No manual rewriting was done — only light human review. The difference is in the process, not in whether an LLM wrote the text. Google has no "this was written by AI" penalty. It has a "does this add value" filter. Content with original data, expertise, and a distinct point of view, accelerated by AI, behaves like any well-crafted page.
Technical myths
Myth 4: the llms.txt file will improve AI visibility. Google has explicitly stated that it does not use llms.txt for content discovery in AI search. A study tracking llms.txt adoption on 10 sites for 90 days before and after implementation found no measurable changes in AI crawl frequency and AI traffic for most sites; the occasional improvement was explained by other SEO efforts. The takeaway, by analogy with sitemaps: llms.txt is infrastructure, not strategy. It won't hurt, but it won't move the needle either.
Myth 5: LLMs handle JavaScript without problems. Research by Profound showed that new AI search providers cannot render JavaScript, and server-side rendering is critical for AI visibility. The myth persists partly because LLMs lie about it. The author asked ChatGPT to summarize a Similarweb article on the impact of AI visibility — a page hidden behind client-side JavaScript. ChatGPT said "no problem" and laid out the key points. Under pressure, it admitted: it had taken them from another readable article. Trust AI answers with caution — especially if your site runs on a headless CMS.
Measurement and people
Myth 6: AI doesn't drive meaningful traffic to websites. The "AI mention effect" study (Profound) showed that AI's impact on traffic is heavily undercounted. After a brand is mentioned in an AI assistant, users visit the brand's site 1.5–2.5 times more often than the predicted baseline within seven days. Even after the May 2026 ChatGPT update, which increased the number of citations, only about 2.5% of those visits carried a trackable AI referral parameter — 97% landed in analytics as direct, organic, or unattributed traffic. Visits are delayed over time: only 20.5% happen within an hour of the mention, 42% within a day, and most first visits occur after the first day. The "AI referrals" segment in your analytics is a floor, not a ceiling.
Myth 7: AI will replace SEO teams. Research by NP Digital (the author is the company's director of SEO) showed that AI finds more errors but misses critical ones and prioritizes poorly. AI catches small errors at scale faster than humans, but evaluates intent poorly when working with keywords and recommends fixing alt text or meta descriptions as the top priority. It also hallucinates fixes, suggests pages that already exist, and confidently agrees with bad ideas. AI is an important tool in the arsenal, but not a full replacement for humans.
Conclusion
Every myth on this list survived because it turned a nuanced topic into a catchy LinkedIn post. AI is changing search fast, and it's hard to keep up with the stream of claims. The only recipe is to slow down, check the data, and be able to accept two facts at once. The winners won't be those who rush to post the loudest take, but those who verified first.
Original: semrush.com
More from Анна Соколова
153 articles
Related articles
- Elvis Presley Song Rights Holders Demand $41.1 Million from Anthropic
- Pinduoduo and Anthropic paid Kazakhstan 8 billion tenge in taxes
- Disney Will Show Super Bowl 2027 Through the Eyes of Star Wars Characters
- Mom's Touch to Open 50 Restaurants in Kazakhstan with Shin-Line
- Prada answered AI slop with a pencil: 6 million views and mice in the style of Disney