swiss made software plus ai & swiss made services plus ai

meeting the needs of swiss tech innovators

The range of technologies that fall under the definition of artificial intelligence - AI - is vast: machine learning, deep learning, predictive maintenance, autonomous systems, computer vision, natural language processing - including large language models and small language models - and generative AI, AI chatbots, AI agents, and more. Software applications that use these technologies, whether as a core function or for specific areas, are constantly growing in number. 

At swiss made software, we want to support Swiss companies that have chosen to embrace these new technologies while staying true to the core values that underpin our “swiss made” label; namely, to certify Swiss origin. With this in mind, we have developed two new variants: the swiss made software + AI and swiss digital services + AI labels.

creating transparency around ai in and for switzerland

The aim is to create transparency for all: This additional label allows Swiss companies to declare their use of AI technology, which itself does not need to be Swiss-made. This additional designation makes it clear to buyers that the software is based on third-party technology and depends on it for core functionality. This goes hand in hand with the recommendation for users to check with the provider to find out what third-party technology or technologies are involved, and how these have been implemented.

how ai is used in swiss it companies

api-based generative ai wrappers – chatgpt, custom gpts, and more

The simplest example is known as a wrapper: An AI wrapper is an application that provides additional functions, often in the form of a user interface or enhanced user experience, which are layered over a generative AI (gen AI) system via an API. These wrappers serve as a bridge between users and the underlying gen AI system, and are intended to improve functionality and user-friendliness. 

If the API connection is interrupted, this service will no longer work. In this case, the key technology is external to the provider. Nevertheless, a service is being provided to make it easier for users to interact with the AI in relation to a specific feature or use case.

generative ai provides additional features

Another variant is the integration of gen AI as a feature: In this case, gen AI is not central to the functioning of the core software. For example, a real estate platform might use generative AI to automatically create texts to describe real estate listings. Regardless of the additional feature, the main function of the software continues to be ability to list and access real estate listings.

open-source and in-house ai solutions, potentially controlled internally

Finally, there are numerous examples where open-source AI or in-house solutions for specific applications are developed by or for companies. It is possible for these to be fully under the control of the company.

It isn’t always easy for users to tell which ways a company is using AI. The swiss made software + AI and swiss digital services + AI labels make it clear that there has been a significant Swiss contribution. At the same time, using one of these labels obliges companies to disclose to their customers exactly where and how AI is being used.

initial questions about the new swiss made plus ai labels:

when and where can the swiss made software plus ai or swiss made services plus ai labels be used?

The new plus AI variant labels can be used wherever AI plays a central role in the offering.

who is eligible for the label?

All companies that produce a significant proportion of their offerings in line with swiss made criteria, but also integrate AI services potentially sourced from third parties.

are there any other requirements companies must fulfil to use the new swiss made plus ai labels?

Companies must place the label on their own website with a backlink to swissmadesoftware.org/plus-AI. The companies must also disclose on their own website how AI is used in their company - whether it is a key technology without which the product or service would not work, or to support the provision of a feature. They must also state whether this is based on third-party technology, open source, or in-house development. 

what should we include when describing an ai offering?

AI as a field is developing rapidly, which makes it difficult to provide a definitive answer. However, we have identified a number of points to make it easier to sum up key areas of interest for users:

  • Hosting: Is it a SaaS model, or is it self-hosted?
  • Level of use and integration: Are you using a basic model, or is it deeply embedded into other services?
  • Type of AI and technologies used: Is this an LLM or generative AI integration, or another type of AI?
  • Origin and type of model: Is it a well-known closed-source model, an open-weight LLM, or open source? Or is it your own proprietary model?
  • Manipulation and fine-tuning: Have existing models been adapted retrospectively?