Switzerland at a crossroads: the Importance of Regulating and Educating on AI

The study conducted in 2023 by IMD, EPFL, and digitalswitzerland, published in January 2024, highlighted a significant adoption of AI in Switzerland, but also a worrying lack of regulatory guidelines and public knowledge. The paper shows that while Swiss organizations are rapidly adopting AI, the population largely remains in the dark about its impacts and potential.

What aspects or trends does the study highlight?

1. Widespread Use of AI: The survey conducted on households and organizations revealed that artificial intelligence (AI) is widely used in both the workplace and personal contexts. The majority of survey participants believe that AI can bring significant benefits to the world. 65% of respondents regularly use AI at home and work, while only 9% state they have had no exposure to AI yet. Additionally, 62% indicate that their organizations have adopted AI, and half of these state that AI is used in three or more business functions, while 31% report that AI is employed in five or more functions.
2. Broad Adoption in Various Sectors: AI is regularly used by people in more than 13 different sectors and industries. Notably, 75% of self-employed workers use AI more frequently compared to 64% of full-time or part-time employees.
3. Positive Attitude but with Reservations: About 84% of survey participants believe that AI can have a positive impact, and 78% are optimistic. 77% foresee that AI will lead to significant changes in the immediate future. However, 32% believe that the potential risks are too high, and 41% are concerned that AI may cause more harm than good.

Is Switzerland an innovative country?

Swiss people are well predisposed towards the widespread use of this technology, and companies seem to be even more so. It is also noteworthy that Switzerland in 2023 once again occupied the first position in the Global Innovation Index (thirteenth consecutive year) ahead of Sweden and the United States, thus it can be considered a nation that highly values innovation, particularly distinguished for creativity, knowledge, technology, and excellent infrastructure.

But then, why is Switzerland taking a cautious approach, with little progress in regulatory and educational aspects?

On the one hand, there is a problem that could be termed cultural: Switzerland often moves slowly compared to other countries in developing norms and directives, leaving the task of self-regulation to companies and individuals. For example, consider anti-corruption laws (on which the OECD has been pressing for years), the protection of whistleblowers in the private sector, the subjection of lawyers and notaries to money laundering laws, privacy regulations, stalking laws, the right to disconnect, etc. However, it is also true that Switzerland is a small country and, as stated by Federal Councillor Albert Rösti in a recent interview on AI published by NZZ am Sonntag, “it is a global issue.” An individual and uncoordinated approach thus seems also senseless and counterproductive even to the head of DATEC.

What are the possible solutions?

The study cited earlier confirms the concerns of the Federal Council and underlines the urgency of regulating a rapidly growing sector. For these reasons, the Federal Department of Communications (DATEC) will present a report to the government by mid-2024, but it has already been specified that there is no intention to hinder innovation in the AI field. One possibility, precisely for the reasons expressed in the previous paragraph, is the adoption of the same rules that will be valid within the European Union. This solution presents undeniable advantages: first of all, it would be simpler and less costly for Swiss companies to adopt the same approach as neighboring countries that represent in many cases a fundamental commercial partner. Moreover, a “sovereign” approach would inevitably prolong the time needed for the implementation and enforcement of binding provisions. By comparison, it took the European Union more than 2 years of work and intense negotiations to arrive at the document that will presumably be approved in April 2024. And it will not yet be the end because, after publication in the Official Journal of the European Union, there will be two years for deferred implementation. Considering the speed with which AI develops and spreads, excessively long times would lead to a risk of “anarchy” difficult to reorder ex-post, especially after companies have taken their own path. It is well known how difficult it is to remodel a system when financial and non-financial investments are particularly significant.

The aspect of regulation represents only a part of the measures to be activated as soon as possible. Perhaps even more relevant is the process of educating the population. When it comes to AI, the vast majority of citizens immediately think of ChatGPT or Gemini, which are only two of the many tools of generative AI and a fraction of the AI world as a whole. A plan of (in)formation and education is therefore urgently needed. More and more people and companies use AI (mainly generative) daily without any awareness of its operation and the risks and issues related to privacy, copyright, and confidentiality clauses, just to name a few.

Moreover, how aware are managers and private citizens that many of the decisions made by algorithms are affected by typically human biases, taken from the data sets on which they have been trained?

How can we protect the rights of citizens, companies, artists when we are faced with “black boxes,” whose outputs are incomprehensible even to their creators?

Are we really able to distinguish what is realistic from what is real, considering the enormous amount of input that we receive 24 hours a day?

To these and other questions, answers are needed. And this is where the concept of critical thinking comes in, an effective tool in protecting us from misinformation, the risk of abuse, etc. Certainly, we are also thinking about the application of digital marks to images, the creation of tools to determine if a text, an image, or a video has been generated by AI (some are already available but with an insufficient degree of reliability), but it is first and foremost we human beings with our intelligence and our thinking who must erect the first barriers useful in protecting us, our children, and our grandchildren from a future with fewer rights and low-quality or polarizing information. Critical thinking arises from knowledge, discussion, comparison and must be trained from an early age. A challenging but more than ever necessary job.

Conclusion

Having recognized the existence of multiple problems, it is now important to find solutions quickly. The federal authorities are working in a certain direction, but we know that the timing of politics is inevitably long, and a regulation on AI will hardly arrive soon. For this reason, it is fundamental that every citizen, regardless of age and position in civil society, carries out a job of self-education but above all changes their approach to everything that today we can read and watch. The creation of realistic content, starting from a simple prompt, is now within everyone’s reach and it is more than ever necessary to evaluate the sources and the veracity of the news before forming an opinion or sharing a news, an image, or a video. Equally relevant is the education of young people, from elementary schools, so that they can develop in them the necessary antibodies to face a world certainly different from the one we have lived so far. Companies also have an important role in all this: they must understand that this technology is certainly a factor of growth, even economic, but this benefit must be shared with their employees, first of all trying to avoid mass layoffs. If only a few people will benefit from the undoubted positive aspects of AI, then we must prepare to face difficult times. Wisdom is what is needed to build a better society. Will we succeed?

Sources

— Report “Charting the Future“;
– Global Innovation Index 2023