In the age of artificial intelligence, this relationship is changing. Google is increasingly incorporating Gemini into its search engine, offering users direct, conversational answers instead of a list of links. As a result, people are staying longer within Google services, and the websites from which the AI draws information are receiving fewer and fewer visitors.<\/p>\n\n\n\n
Google has turned traditional search into a conversation with artificial intelligence with \u201cAI Mode.\u201d Users can ask longer questions, add photos and videos, and let AI agents perform more complex searches.<\/p>\n\n\n\n
According to the company, queries in AI Mode are about three times longer than traditional search strings, and research shows that users spend between one and nine minutes more in the new mode than with regular searches.<\/p>\n\n\n\n
A study by Growth Memo found that users did not click through to any external websites in about 75 percent of the sessions analyzed. Google rejects the research methodology, but publishers and other companies are already reporting a significant decline in search traffic.<\/p>\n\n\n\n
Nilay Patel, editor-in-chief of technology publication The Verge, has for years described this development with the term \u201cGoogle Zero\u201d \u2013 the moment when the traffic publishers receive from Google approaches zero. In his opinion, this moment has already arrived for many media outlets.<\/p>\n\n\n\n
The classic search engine acted as a gateway to the open web. It offered links to the user, but they had to find the answer on the websites of media, companies, universities, and other creators.<\/p>\n\n\n\n
Generative search turns this model on its head. The AI scans the content on the web, compiles an answer from it, and presents it directly to the user. The user often no longer has a reason to visit the original page.<\/p>\n\n\n\n
For Google, this development is beneficial, as users spend more time in its ecosystem. For websites, it means fewer visitors, less advertising revenue, and less direct relationships with their audience.<\/p>\n\n\n\n
Cloudflare data shows that more than half of all web traffic is already generated by automated systems. Between June 2025 and April 2026, human traffic to websites in sectors such as media, finance, and commerce is expected to decline by almost 40 percent.<\/p>\n\n\n\n
Wikipedia is also feeling the changes. The Wikimedia Foundation has seen an eight percent decline in human visitors over the past year, while the number of robots that extract data from the encyclopedia to train artificial intelligence models has increased.<\/p>\n\n\n\n
Wikipedia is therefore looking for new ways to connect directly with readers, including through its own app and social media. It has also started charging Google and other AI companies for commercial access to its data.<\/p>\n\n\n\n
The situation reveals a fundamental problem with the new online economy: Websites fund the creation and curation of information, and AI companies can use that information to deliver answers that never lead users to the original pages.<\/p>\n\n\n\n
If content creators lose their audience and revenue, the amount of quality information on which search engines and AI models are based will also decrease in the long term.<\/p>\n\n\n\n
Google claims it is not threatening the open web and that its AUI features generate billions of clicks to external pages every week. The company has introduced site previews, the ability to choose preferred sources, and profiles that allow publishers and creators to better showcase their work.<\/p>\n\n\n\n
According to Google, the overall number of clicks remains relatively stable, but critical research is said to use flawed methodologies. The company also emphasizes that answers in AI Mode include links to original content.<\/p>\n\n\n\n
But publishers' experience paints a different picture. Vox Media, the owner of publications like The Verge, has cited declining search traffic as one of the reasons it is looking for new business models and selling off parts of its websites.<\/p>\n\n\n\n
Regulators are already getting involved in the debate. The UK's competition authority has ordered Google to more clearly attribute sources in AI responses and make it easier for users to access the original web pages.<\/p>\n\n\n\n
Page owners should also be given the option to exclude their content from UI summaries. Google is expected to roll out some of these changes more broadly, not just in the UK.<\/p>\n\n\n\n
For regulators, the key question is whether Google can use open web content to keep users in its own environment without providing websites with adequate traffic or other compensation.<\/p>\n\n\n\n
The World Wide Web was conceived as an open space where anyone can publish content and users can move freely between different pages. Social networks and mobile platforms have already locked much of this traffic into their own ecosystems, and artificial intelligence is accelerating the process.<\/p>\n\n\n\n
Google remains more open than networks that require login to access. The user can still leave the search engine and visit the original website. But increasingly, they simply don't need to.<\/p>\n\n\n\n
This is where the biggest risk lies. Google may not be closing the open web directly, but it is reducing the economic value of visiting the sites from which it draws information.<\/p>\n\n\n\n
If the web becomes primarily a database for AI responses, its business model will also have to change. Otherwise, AI could exhaust the very ecosystem on which it depends.<\/p>","protected":false},"excerpt":{"rendered":"
V obdobju umetne inteligence se ta odnos spreminja. Google v iskalnik vse mo\u010dneje vklju\u010duje Gemini in uporabnikom namesto seznama povezav ponuja neposredne, pogovorno oblikovane odgovore. Posledica je, da ljudje dlje ostajajo znotraj Googlovih storitev, spletne strani, iz katerih umetna inteligenca \u010drpa informacije, pa prejemajo vse manj obiskovalcev. Uporabniki vse redkeje zapustijo Google Google je z […]<\/p>","protected":false},"author":2,"featured_media":0,"comment_status":"","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[4],"tags":[126,192],"class_list":["post-7851","post","type-post","status-publish","format-standard","hentry","category-racunalnistvo-telefonija","tag-google","tag-umetna-inteligenca"],"acf":{"subtitle":"Ko je Google leta 2004 vstopal na borzo, je njegov soustanovitelj Larry Page zapisal, da mora imeti dobro delujo\u010da dru\u017eba prost in nepristranski dostop do kakovostnih informacij. Google je to poslanstvo uresni\u010deval tako, da je uporabnike prek rezultatov iskanja usmerjal na milijone spletnih strani.","heading":"","summary":"Ko je Google leta 2004 vstopal na borzo, je njegov soustanovitelj Larry Page zapisal, da mora imeti dobro delujo\u010da dru\u017eba prost in nepristranski dostop do kakovostnih informacij. Google je to poslanstvo uresni\u010deval tako, da je uporabnike prek rezultatov iskanja usmerjal na milijone spletnih strani.","thumbnail_small":"https:\/\/racunalniske-novice.com\/wp-content\/uploads\/2026\/05\/gemini-3-5__keyword__blog-header.width-1200.format-webp-560x315.webp","thumbnail_large":"https:\/\/racunalniske-novice.com\/wp-content\/uploads\/2026\/05\/gemini-3-5__keyword__blog-header.width-1200.format-webp-1024x577.webp","thumbnail_caption":"","gallery":"","video_gallery":null,"author":"","links":null,"sources":[{"title":"NYTimes","url":"https:\/\/www.nytimes.com\/2026\/07\/20\/technology\/google-ai-open-web.html?unlocked_article_code=1.zFA.7A4y.kcG9Q_Hnu0C-&smid=bs-share"}],"skip_language":[]},"yoast_head":"\n