Google AI mode for Switzerland - search now thinks for itself

From the Link to the Answer: A New Search Experience

On October 8, 2025, Google officially launched the new AI Mode in Google Search in Switzerland. This is currently being rolled out in almost 50 additional countries. It may therefore take a few days for the new tab with AI mode to appear on every device. AI mode appears as an additional tab in Google Search and in the Google app for Android and iOS.

In contrast to the classic search, which provides hit lists in the form of "blue links", the AI mode presents directly generated answers in continuous text form. This turns the search into a dialog: Users can ask multi-layered questions that would previously have required several searches - for example, when exploring a topic, comparing options or planning a trip.

Users can now ask questions such as:

«I want to understand the different methods of making coffee. Please create a table for me that compares the differences in terms of taste, ease of use, and required equipment.»

And even follow up with a follow-up question, such as:

«What is the best grind size for each method?»

Initial tests show that search queries in AI mode are two to three times longer than conventional queries. After the first answer, the dialog can be continued - with follow-up questions that are interpreted in the context of the conversation. The search experience thus resembles a natural conversation with an AI assistant.

Technology with Depth: «Query Fan-Out» and Gemini Power

The technological basis is an adapted version of the Gemini models (Gemini 2.0/2.5), which supports advanced logical reasoning (Advanced Reasoning). Central to this is the new «Query Fan-Out» technique: It breaks down a search query into several sub-questions, researches these in parallel in different data sources - including the search index, the knowledge graph and the shopping graph - and combines the results into a coherent answer. The AI mode is also multimodal: queries can be made by text, voice or via the camera.

Quality, Safety, and Transparency

Google is integrating AI mode into its Quality and Evaluation Systems. New methods are being added to improve factual accuracy and certainty. For example, agentic reinforcement learning (RL) is used to reward models for generating verifiable and precise statements.

Known protection mechanisms such as SafeSearch and SpamBrain remain active. Nevertheless, Google emphasizes that the AI mode is not infallible: Like any AI system, it can misinterpret content or, in rare cases "hallucinate" - i.e. present fictitious information with confidence. Users are therefore asked to check important information with several sources. Google continues to provide source links to ensure traceability.

Control is also ensured when it comes to data protection: personalized results appear only if the «Web & App Activity» feature is enabled. You can view and delete your search history at any time.

SEO in Transition: From Keywords to Entities

With the introduction of generative search experiences, the focus in search engine optimization is shifting: entities are replacing keywords.

What are entities? In the SEO context, an entity refers to a clearly identifiable object - a person, a place, a thing, an idea or a concept - defined by name, attributes and relationships to other entities.

The role of the Knowledge Graph: An entity only "exists" online if it is listed in the Google Knowledge Graph. This was introduced in 2012 with the principle "Things not strings" and forms a network of semantically linked topics. This means that Google increasingly understands queries contextually rather than linguistically - even without exact keyword mentions.

Strategic Implications: Because the AI mode delivers answers directly, a decrease in organic search traffic is expected. It will be crucial to be cited in AI-generated answers.

In the future, websites will need to structure their content in such a way that AI systems can interpret it unambiguously:

1. Structured data: Markup codes in the head section help clearly define entities and their relationships. Tags such as sameAs, About or mentions provide Google with semantic cues.

2. Content Focus: Instead of counting keywords high-quality, relevant content related to entities. Even niche topics are gaining traction, since 15 % of daily search queries are completely new.

3. Internal links: Through Pillar-Based Marketing and strategic internal links can strengthen relationships between entities.

From Search Index to Knowledge Network

The AI mode does not mark an evolutionary step, but a revolutionary one: Google search is evolving from an index to an answer engine - and from keyword hunting to context intelligence. Traditional search results retain their relevance - for example for transactional queries or queries that need to be checked. However, the new AI mode clearly shows that the future of search is dialogical, entity-based and generative.