Europe is moving to regulate artificial intelligence (AI), tighten security standards for chatbots such as ChatGPT, and impose a regulatory framework on developers and manufacturers who want to do business in the EU internal market – an ambitious challenge on the global market, where the United States and China simply let technological progress run wild. The European Commission proposed a new set of rules two years ago, the so-called Artificial Intelligence Act, which is currently being negotiated in Brussels and is set to be the first-ever legal framework on AI.

On 28 April, the political groups in the European Parliament reached a provisional agreement on the EU’s AI rulebook, which mainly focuses on so-called “foundation models”, primary knowledge building blocks that serve as the fundament for AI technologies. MEPs want these elements to be rigorously “designed and developed in accordance with EU law and fundamental rights, including freedom of expression.”

The text based on the informal agreement will now have to go to the European Parliament’s Internal Market Committee for a vote on May 11. A plenary vote on the European Parliament’s negotiating position is expected in June, followed by negotiations with the Council and the Commission.

Endless possibilities versus unpredictable risks

Sabine Bendiek, Head of Human Resources at German software company SAP, said that AI could enormously improve productivity and be a support for people. For example, extremely monotonous, repetitive tasks could be taken over by AI. “Our employees can then really focus on using what makes humans so strong: creativity and the ability to evaluate results with a different perspective and implement them accordingly.”

But what about current and future developments based on algorithms and artificial intelligence? What about ChatGTP and autonomous cars? What about the risk that AI could be used to spread disinformation? Several MEPs recently called for a global summit on the dangers of artificial intelligence.

Last month, German Digital Minister Volker Wissing called for the swift creation of a legal framework for AI in Europe. “We must react wisely now and regulate artificial intelligence sensibly before it is too late for that. This must not take years again,” Wissing told the German newspaper Bild am Sonntag. A spokeswoman for the German Interior Ministry told the daily Handelsblatt that “it is essential to strike a balance between openness to innovation and a clear legal framework that defines standards for trustworthy AI.”

“Artificial intelligence regulation is at a very low level and its acceleration must be maximized,” said Dobroslav Dimitrov, Chairman of the Bulgarian Association of Software Companies (BASSCOM) and Bulgarian Employers’ Association of Innovative Technologies (BRAIT). In his opinion, the initiative is definitely overdue, which is why all decision makers at government, corporate or educational level should look into what is going on and adjust themselves on the move.

Dimitrov compared AI with nuclear energy. “It can set the whole planet in motion, but it can also destroy it if used for destructive purposes. The same applies to artificial intelligence,” he said.

Italian Minister for Economic Development Adolfo Urso stated that it was necessary to develop new regulations to better define the relationship between man and machines, as machines seemed to have the advantage for the first time.

ChatGPT under scrutiny, causes acceleration of AI legislation

Recalling the case of AI chatbot ChatGPT, which Italy had temporarily banned in March and made available again at the end of April, Urso stressed that it “showed the pitfalls of a sector that still needs strict regulation.” He added that the Italian data protection authority’s provisional restriction of ChatGPT “has resulted in a higher level of protection for users.”

Early last month, French Digital Minister Jean-Noël Barrot said that ChatGPT did not comply with the EU General Data Protection Directive (GDPR), but that it was better to “frame” it rather than ban it. In mid-April, the French National Commission on Informatics and Liberty (CNIL) decided to open an investigation into five complaints against the chatbot.

The European Data Protection Board (EDPB), which is the European level equivalent of the CNIL in the various member states and responsible for coordinating the authorities, announced the creation of a dedicated working group to promote European cooperation on the subject.

Applying AI in the workforce

Urso recently stated that the artificial intelligence sector in Italy, for example, had grown by 32 percent in one year to a value of 500 million euros. The sector “is now central to the development programs of companies of all sizes that invest in intelligent data processing, language AI and computer vision,” he said.

In the general business landscape in Spain, artificial intelligence is used in 8 percent of companies with more than 10 employees. According to data for 2021 published by the EU statistics office Eurostat, this figure corresponds to the European average. Production processes, marketing and sales strategies, cybersecurity and the organisation of management processes are the main applications of AI in Spanish companies, according to a 2022 study by the Spanish multinational financial service company Banco Bilbao Vizcaya Argentaria (BBVA).

The Croatian Association for Artificial Intelligence (CroAI) gathers leading companies and startups in the field of artificial intelligence in Croatia. According to data from that association, more than 130 startups in Croatia were developing and looking for the application of artificial intelligence as of the end of 2022. While three years ago, 220 companies were applying AI, that number had grown to around 440 by October 2022, CroAI said.

Europe as a global player in AI

Bulgarian Interim Minister for Economy and Innovation Alexander Pulev recently noted that Bulgaria could become a regional hub for the flow of digital information from Asia. Furthermore, he said that Bulgaria and the US government had signed a declaration aimed at improving the quality of Bulgaria’s digital infrastructure with a focus on cyber security.

The President of the Foundation “Society of Common Values” in Bosnia and Herzegovina (BiH), Zlatko Lagumdžija, pointed out the importance of integrating the digital transformation of the EU, which should not be limited to EU member states, but should include neighbouring countries like those of the Western Balkans on the same principles and projects.

Speaking at a conference in Cambridge, Lagumdžija noted that only in this way could the EU demonstrate that it had the capacity to lead these processes on a global level, together with, for example, the United States and Japan.

The EU was already deeply into the process of creating an online regulatory framework, and must go back to the drawing board now to figure out how to effectively regulate AI.The battle over an outright ban on the use of artificial intelligence facial recognition technologies without the consent of those involved, where the EU Council could push through an exception related to national security clauses, is expected to be particularly bitter.

This article is published Fridays. The content is based on news by agencies participating in the enr.