Submitted text
The Federal Council is instructed, Measures for the creation of a competence center for new technologies and in particular for artificial intelligence (AI) in Switzerland. This center should contribute in particular to achieving the following goals:
- Switzerland has a competence center that can advise the authorities and possibly the private sector;
- Switzerland has the computing power and infrastructure for the development, testing and research of AI;
- Switzerland has tools and can provide tools that make it possible to deal with artificial intelligence, e.g. by recognizing risks, preventing them or helping to better understand them;
- Switzerland is constantly monitoring developments to ensure that it does not react too late in terms of regulation and development;
- AI is promoted in the interests of the common good and the population and these technologies are prevented from developing to the detriment of the vast majority of the population.
The creation of such a center can take place in various ways, e.g. by:
- the universities’ budgets are linked to inflation, allowing them to provide resources for the development of or contribution to such a center;
- playing a facilitating role in the development of such a center by providing resources and creating public service contracts for this center;
- a public mandate is defined and put out to tender that ensures that AI can develop in Switzerland, but benefits everyone and is as transparent as possible;
- Funding or co-funding the development of computing power that can be made available; this aspect is indeed fundamental to research, testing and deployment in the field of AI;
- training courses, thanks to which Switzerland can position itself as a leader in this field.
Justification
Technologies are developing very quickly, which is not always compatible with a certain inertia in the political system. This can lead to the legal system no longer keeping pace with developments. Switzerland should play a pioneering role in new technologies, but always keep the interests of society at heart. Switzerland should therefore be a country where the development of technologies is encouraged while ensuring that these technologies benefit humanity, rather than serving the interests of a few individuals or companies and massively weakening others.
A competence center would enable Switzerland to play a pioneering role and constantly monitor the development of technologies, particularly in the field of AI. Such a center could help policymakers to react quickly and would help to ensure that AI technologies are transparent. This could limit risks and procure or develop tools that increase the value of these technologies for the population.
The skills are already very much in place in Switzerland, but there is a lack of political impetus and resources. If universities and other stakeholders were encouraged to develop such a center, synergies with the private sector could be created that would position our country as a leader in the field of AI for the common good and create an environment conducive to business.
Switzerland has a serious lack of resources, i.e. available digital infrastructures, in particular available computing power, to enable institutions and partners to research, develop, test and learn in the field of AI. It would be very important to substantially increase these capacities so as not to be too dependent on other countries.
Opinion of the Federal Council of 30.8.2023
AI raises questions in the most diverse areas of our society and economy, whereby the effects of AI are highly context-dependent. Against this backdrop, a single competence center responsible for all technical, legal and social aspects of AI at sectoral and horizontal level appears neither expedient nor feasible. The federal government has already created several bodies to exchange information on various aspects of AI. Since 2019, the Tripartite Platform coordinated by OFCOM has been used to discuss political, social and economic issues relating to AI with over 300 experts from the administration (Confederation and cantons), business, civil society and academia. The Plateforme Tripartite is open to all interested parties and has an administrative committee to coordinate Swiss positions on AI. For several years, AI has also been a priority topic at the Swiss Internet Governance Forum, which is under the patronage of OFCOM.
Since 2021, the Competence Network AI (CNAI) is responsible for technical aspects of the specific use of AI in the Federal Administration. The CNAI, which is affiliated to the Federal Statistical Office (FSO), maintains a database of AI projects in the Confederation, coordinates AI-related terminology and networks technical experts in the “Community of Expertise” and “Community of Practice”. A “legal hub” is currently being created to support the CNAI and the Plateforme Tripartite management committee in legal matters. Since 2021, there has been an “Algorithms Node” under the leadership of the FSO’s Data Science Competence Center (DSCC). The Federal Council has instructed the DHA to evaluate the CNAI by the first quarter of 2024 and, if necessary, to propose further developments. In addition, the DSCC provides the public sector with its expertise in the specific application of data science and AI. In addition, OFCOM, in consultation with the Management Committee of the Plateforme Tripartite, conducts a biennial monitoring of the use and timeliness of the 2020 AI guidelines for the Confederation and collects the planned measures at sectoral level. Finally, the DDPS technology center, armasuisse Science and Technology, conducts research into AI applications in the security sector.
Also on various universities competence centers for AI have been created, which can be used for the concerns mentioned in the motion. This also applies to computing power and infrastructures. The new Swiss supercomputer “alps”, which is currently being procured by ETH Zurich to replace the existing “Piz Daint” computer, will once again be one of the most powerful supercomputers in the world. The supercomputer should be ready for use from Q1/2024.
The Federal Council is convinced that dealing with AI (and other digitalization issues) will continue to gain in importance. Whether this will require an adaptation or strengthening of the existing institutional structures in the field of AI in the future also depends heavily on regulatory developments at international level. In addition to the EU’s negotiations on the “AI Act”, the negotiations in the Council of Europe under the Swiss Presidency for a binding international agreement on AI, human rights, democracy and the rule of law are particularly relevant. In its response to the Postulate Dobler (23.3201) has already been announced by the Federal Council, by the end of 2024, a political assessment with possible options for sectoral and, where necessary, horizontal regulatory measures in the area of AI. The analyses for this will be carried out within the framework of existing bodies (in particular the IC EUDP for the impact of the AI Act on Switzerland, the Tripartite Platform and its management committee for the analysis of the implications of the Council of Europe Agreement, the monitoring of the guidelines and the exchange on the need for sectoral action, as well as the CNAI) and with the involvement of all federal agencies concerned. The analysis should also examine the question of whether the existing institutional structures in the area of AI should be adapted or strengthened (e.g. creation of an overarching coordination office).