Legal Tech
ChatGPT vs Silex: Which AI Is Better for Legal Research in Switzerland?

Artificial intelligence is increasingly used by legal professionals to save time and improve productivity. Tools like ChatGPT make it easy to generate text, summarize documents, and explore legal topics in seconds. However, when it comes to legal research, accuracy, reliable sources, and jurisdiction-specific knowledge are essential.
This is particularly true for Swiss law, where legislation and case law are complex and multilingual. While general AI tools can be useful, they are not designed specifically for legal reasoning or Swiss legal sources.
In this article, we compare ChatGPT and Silex, a legal AI built specifically for Swiss law, to see which tool is better suited for legal research.
Comparing ChatGPT and Silex for Legal Research
While both ChatGPT and Silex use artificial intelligence to answer questions, they are designed for very different purposes. ChatGPT is a general-purpose AI assistant, while Silex is a specialized legal AI built for Swiss law. This fundamental difference affects how each tool performs when used for legal research.
Feature | ChatGPT | Silex |
Purpose | General-purpose AI used for writing, brainstorming, summarizing documents, and answering broad questions across many domains. | AI designed specifically for legal professionals and legal research in Swiss law. |
Sources | Trained on large amounts of general internet data and does not systematically reference legal databases. | Built on Swiss legal sources, including legislation and case law. |
Reliability | Can sometimes generate hallucinations, meaning confident answers that contain incorrect or invented legal information. | Designed to reduce hallucinations by relying on structured legal data and verified sources. |
Legal methodology | Generates text based on language patterns and probabilities rather than legal reasoning. | Built to support legal analysis and structured legal reasoning. |
Confidentiality | Not specifically designed for the strict confidentiality requirements of legal practice. | Emphasizes privacy and security with Swiss hosting and no training on user data. |
Purpose
ChatGPT is designed to assist with a wide range of tasks, such as writing, brainstorming, summarizing documents, or answering general questions. It is trained on large amounts of publicly available data and aims to generate fluent and helpful responses across many domains.
Silex, on the other hand, is specifically designed for legal professionals. Developed by the Swiss legal tech startup Ex Nunc Intelligence, it focuses exclusively on legal research and legal reasoning within the Swiss legal system.
Sources
One of the main differences lies in the sources used to generate answers.
ChatGPT relies on broad training data and does not systematically reference specific legal databases. As a result, it may provide incomplete or inaccurate legal information, especially for jurisdiction-specific questions.
Silex is built on legal sources related to Swiss law, including legislation and case law. Its answers rely on documented legal materials rather than general knowledge.
Reliability
Because ChatGPT generates responses based on statistical patterns in language, it can sometimes produce hallucinations, meaning confident answers that contain incorrect or invented information.
Silex aims to reduce this risk by focusing on structured legal data and verified legal sources, allowing users to work with information that better aligns with professional legal standards.
Confidentiality
Another important consideration for legal professionals is data confidentiality.
ChatGPT is not specifically designed for the strict confidentiality requirements of legal practice. By contrast, Silex emphasizes privacy and security, with Swiss hosting and a policy of not training its models on user data, helping preserve professional confidentiality.
Why Generic AI Falls Short for Swiss Legal Research
General-purpose AI tools can be useful for drafting text or exploring broad topics. However, legal research requires a much higher level of precision and reliability, especially in a jurisdiction such as Switzerland.
Swiss law is complex and highly structured. It includes federal legislation, cantonal regulations, and extensive case law, often written in multiple languages. Legal professionals must rely on accurate sources, verifiable references, and clear legal reasoning when answering a legal question or preparing a case.
Generic AI tools are not specifically designed to handle this type of legal environment. Because they rely on broad training data rather than structured legal sources, they may produce answers that appear convincing but lack proper legal grounding. In some cases, they may even generate references to legal provisions or court decisions that do not exist.
For lawyers, notaries, and legal departments, this creates an obvious risk. Legal research is not just about obtaining an answer quickly, it is about being able to verify and trust the sources behind that answer.
Why Silex Is Designed for Legal Research
Silex was developed specifically to address these challenges. Rather than acting as a general chatbot, it is designed as a legal research assistant built for the Swiss legal system.
The platform relies on structured legal sources, including legislation and case law, allowing users to work with answers grounded in documented legal materials. Its design also follows a methodology closer to the way lawyers conduct legal research in practice.
For legal professionals, this approach provides several advantages:
Faster access to relevant legal provisions and case law
Reduced time spent on manual research
Answers supported by identifiable legal sources
A workflow aligned with legal reasoning
Instead of replacing lawyers, tools like Silex aim to augment their work by making legal research faster, clearer, and more reliable.

