AI assistants challenge Google’s search dominance as users shift to LLMs
Growing numbers of people use large language models like ChatGPT for everyday queries and planning, prompting changes in search, marketing and verification practices
A growing number of internet users are turning to large language models (LLMs) such as OpenAI’s ChatGPT for everyday queries and recommendations, posing a significant challenge to traditional search engines including Google, industry researchers and users say.
Users describe LLMs as a faster, lower-effort way to get concise answers and personalised suggestions. "I've always been an early adopter… and in the past year have started using ChatGPT for just about everything. It's become a second assistant," said Anja‑Sara Lahady, a lawyer and legal technology consultant in Montreal, who uses LLMs for low‑risk professional tasks and everyday decisions such as meal ideas and travel planning.
Usage figures indicate rapid growth. Demandsage, a data and research firm, estimates ChatGPT attracts around 800 million weekly active users, up from 400 million in February 2025. Research firm Datos reported that 5.99% of desktop browser searches in July went to LLMs, more than double the share a year earlier. Despite that rise, analysts say LLM use is still a minority behaviour compared with traditional search.
Academics and industry experts say people are drawn to LLMs because they reduce "cognitive load" by synthesising multiple sources into a short, editable answer. "Instead of juggling 10 links with search, you get a brief synthesis that you can edit and iterate in plain English," said Feng Li, associate dean for research and innovation at Bayes Business School in London. He added that LLMs are particularly useful for summarising long documents, first‑pass drafting, coding snippets and exploratory "what‑if" queries, but warned outputs still require verification because hallucinations and factual errors remain common.
Google has resisted claims that LLMs are eroding its search business. The company said overall and commercial queries continued to grow year‑over‑year and credited new AI features, including AI Mode and AI Overviews, with improving the search experience. The introduction of those tools lets users ask more conversational questions and receive tailored summaries at the top of the search results page.
However, evidence of an impact on Google emerged during testimony in a US Department of Justice antitrust trial. A senior Apple executive said the number of Google searches on Apple devices via the Safari browser fell for the first time in more than 20 years.
The rise of LLMs is prompting companies to rethink marketing strategies. "They need to understand which sources the model considers authoritative within their category," said Leila Seith Hassan, chief data officer at Digitas UK, citing differences in the sources LLMs reference across markets. She said LLMs often place more trust in official websites, press releases, established media and recognised industry rankings than in social media, and that referrals from LLMs can produce higher‑quality leads with better conversion rates.
Media and marketing professionals report practical benefits from using models for time‑consuming research. Hannah Cooke, head of client strategy at Charlie Oscar, said she uses ChatGPT and other models to streamline tasks such as personalised skincare recommendations and travel planning, noting the tools can save hours of research by filtering options and suggesting suitable restaurants or itineraries.
Experts predict a hybrid future rather than a wholesale replacement of search. Professor Li said LLM usage will likely stabilise as people reserve models for creative, summarising and exploratory tasks while continuing to use traditional search for transactions, bookings and verification. That pattern would leave commercial search functions — shopping, booking, transaction-oriented queries — largely with established engines.
Companies that rely on search visibility and marketers seeking to drive conversions face an evolving landscape: they must not only maintain authoritative, high‑quality online content but also understand how LLMs surface and attribute information. At the same time, users and businesses are being reminded to verify model outputs, particularly for high‑risk or legally significant decisions.
As LLMs become more capable and more widely used, technology companies, regulators and businesses will continue to monitor how conversational AI reshapes search behaviour, commercial models and the mechanisms users rely on for accurate information.