The future of search - how AI is reinventing search engines
From search engine to answer engine: Companies need to rethink their strategy today. For decades, «Googling» was synonymous with searching for information. But search behavior is changing rapidly.
Artificial intelligence is penetrating deep into the core function of search engines and reshaping them. Google is rolling out its AI Mode, ChatGPT, Perplexity and Grok are taking on the role of digital advisors, and the first AI agents are already acting independently. The question is therefore no longer: Do we still need traditional search engines? But rather: In what form will we use them in the future - and who will control the answers?
Status quo: change is already a reality
The upheaval is no longer a future scenario - it is happening now. In March 2025, 13 % of all Google searches already contained an AI overview, an AI-generated answer block above the organic results. Just a few months earlier, this figure had been just six percent - a doubling in record time.
These overviews mainly appear for informational search queries, but are increasingly also penetrating commercial and navigation-related search intentions. A study by Pew Research shows that users only click on traditional results in 8 % of cases when an AI summary is displayed (without summary: 15 %). Within the AI answers, only around 1 % of users follow the linked sources.
Publishers are reporting massive losses in reach. In some cases, the number of clicks dropped by up to 80 %. Google itself is also changing: according to Semrush, only 6-8 % of sessions led to external clicks in the first few months after the introduction of AI Mode. Over 90 % of users therefore remained in the Google universe.
The relationship between query and evaluation is thus fundamentally shifting. Brands, publishers and agencies need to rethink their understanding of visibility.
How do users search today and tomorrow?
People search differently today. Instead of short keywords, they formulate entire questions such as "Which espresso machine suits my lifestyle?". They expect a precise answer, directly in the interface, without further scrolling.
The search becomes more dialogic. Users ask questions, refine topics, switch between chat and web search - depending on the trust and complexity of the query. There are also new modalities: voice input, image recognition, AR glasses and smart devices that display information contextually. Anyone looking at a building, for example, immediately receives historical data, ratings or prices.
An international study with over 12,000 search queries shows: Users are more skeptical of AI answers than traditional results. However, as soon as sources are visible, trust increases significantly. AI must therefore not only provide answers, but also explain where it gets its knowledge from.
In short: the search of the future is multimodal, dialogic and context-aware. Users don't want lists of links, they want comprehensible, trustworthy answers.
Challenges and risks
As the new possibilities increase, so do the risks.
A core problem is the question of sources. Which content is cited - and which is not? Studies show that AI Overviews are heavily dominated by just a few platforms. Wikipedia, Reddit and Quora appear particularly often. According to Semrush, just 50 domains account for 28.9 % of all AIO mentions. This means that a few websites determine which information remains visible. While many publishers lose clicks, these top platforms may even benefit. For smaller providers, visibility thus becomes a luxury good.
The susceptibility of AI to errors also remains an issue. Systems can invent facts or take statements out of context. Such hallucinations not only jeopardize trust in the technology, but also the reputation of the brands mentioned. Monetization is also still unresolved. Advertising in AI responses is still at the experimental stage. How can ads be integrated without undermining credibility? The balance between relevance and manipulation is fragile. There are also legal issues. European publishers have already filed complaints against Google, and companies such as Chegg are suing for unpaid use of their content. At the same time, the industry is discussing new standards to regulate AI crawlers and protect content rights.
Finally, there is a risk of bias effects: AI systems reproduce existing biases. If only dominant perspectives are reinforced, diversity of opinion dwindles. Trust only arises when users feel that answers are fair, comprehensible and balanced.
Recommendations for companies and agencies
The good news is that those who act now can benefit from the change.
- Optimize content for AI: Content must be structured in such a way that it can be recognized and cited by AI systems - clear structure, understandable language, clean data.
- Transparency and source competence: Show where your information comes from. Precise source details increase trust and the chance of appearing as a reliable reference in AI summaries.
- New performance measurement: In future, it will no longer be the click that counts, but the impact. Micro-conversions, dwell time and engagement are the relevant KPIs.
- Experiment: Test how your content is displayed in AI Overviews. Optimize the language, layout and data structure on an ongoing basis.
- Hybrid strategies: Classic SEO remains important, but a bridge to generative search is needed. Visibility in both worlds determines future viability.
In the end, it's all about trust and quality. Companies that communicate openly, precisely and consistently will not be ousted in the new search world - they will be strengthened.
Where the journey leads
A look into the near future shows that search will change more than we realize today. Three development paths are emerging.
- Search as agent
In this world, the AI not only responds, it acts. A sentence like "Plan me a weekend in Rome" is enough - the assistant books flights, hotels and restaurants automatically. Convenience replaces control. The question is no longer which information is correct, but who we trust to act on our behalf.
- Invisible search
In another future, search disappears completely from our consciousness. AR glasses and wearables deliver information directly into our field of vision. AI knows what we need even before we ask. This is convenient, but also unsettling - because the line between assistance and influence is becoming blurred.
- Own search engines
The counter-model: decentralization. Users operate their own AI systems, select sources themselves and share models in small communities. This creates a personal search that reflects values and perspectives. But here too, a new form of filter bubble is looming. It remains to be seen whether one of these scenarios will become reality or whether all three will merge. One thing is certain: search is becoming more personal, more ubiquitous and more powerful. Anyone still clinging to traditional SEO thought patterns will be left behind.
Why now is the time to rethink
We are on the threshold of a completely new understanding of search. The classic search engine remains, but it is transforming into an answer engine that understands contexts, conducts conversations and prepares decisions. Search is changing from a tool to a digital partner.
For companies, this means a paradigm shift. Keywords and rankings are no longer enough. The decisive factor is whether a brand appears in the AI's answers - and whether it sounds credible there. Those who invest in response quality, source competence and strategic AI optimization today will secure visibility and trust tomorrow.
Because that's what it's all about: Users want answers that are well-founded, transparent and relevant. Brands that offer this will survive in a world where search engines become assistants and conversation partners. The future of search is not a distant trend. It is the beginning of a new dialog between people, brands and machines - and it has already begun.
Sources:
https:/ www.semrush.com/blog/semrush-ai-overviews-study/
https:/ www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely to-click-on-links-when-an-ai-summary-appears-in-the-results/ https:/ www.semrush.com/blog/google-ai-mode-seo-impact/
https:/ arxiv.org/abs/2504.05156
https:/ arxiv.org/abs/2504.06435
https:/ www.theguardian.com/technology/2025/jul/24/ai-summaries-causing-dev astating-drop-in-online-news-audiences-study-finds
https:/ www.reuters.com/legal/litigation/googles-ai-overviews-hit-by-eu-antitrust complaint-independent-publishers-2025-07-04
https:/ www.theverge.com/news/619051/chegg-google-ai-overviews-monopoly
