The use of AI-driven applications is steadily increasing, exposing a number of problem areas associated with this technology.
The EU Commission is presenting a draft law that aims in particular to prevent the discrimination and surveillance of humans by artificial intelligence (AI).
A uniform, EU-wide legal framework is to be created to which all users and developers of artificial intelligence must adhere.
It is difficult to overlook the benefits of artificial intelligence in countless fields of application, especially for companies. Nevertheless, this key technology brings with it a number of problems, which Europe-wide regulation will seek to solve. Users and developers of artificial intelligence can expect a comprehensive regulatory framework.
Previous freedoms in dealing with artificial intelligence
Artificial intelligence (AI) is the key technology of the current era. As this technology has rapidly developed, numerous applications have been formed that are based on it. As a result, companies have been able to benefit in many areas from enormous increases in efficiency and cost reductions, and also with improved output. Existing systems have thus been optimised and new business areas opened up.
Under the leadership of the EU Commissioner for Competition, Margrethe Vestager, the EU had previously given particular support to the developers of such applications. The goal was to remain globally competitive as an innovation location for artificial intelligence. The previous drivers of this technology, the USA and China, seemed to be gaining ground in terms of "digitalisation through AI", especially in the last decade.
As a result, numerous lighthouse projects have emerged in the EU, primarily in the areas of smart cities, smart energy and the health sector. Self-driving cars, intelligent industrial and energy systems and automated drug development are just some of the significant areas that will determine life in the future and can only be realised through the use of AI-driven software.
The dark side of artificial intelligence
In addition to the lighthouse projects within the EU described above, artificial intelligence can and is also used to analyse and evaluate people. These applications are bundled under the term "social scoring". Users of such applications use data obtained, e.g. in the form of application documents, to select an optimal application candidate on the basis of formal data. Similarly, banks use existing loan repayments to analyse the extent to which a new prospective customer would be creditworthy on the basis of his or her data.
The use of these fields of application of artificial intelligence thus leads directly to unequal treatment of individuals and creates a social gap and social inequality that contradicts the principles of the EU and its member states. At this point, it quickly becomes clear that unconditional support from the EU does not contribute to the desired goal in all use cases.
EU regulation as the optimal solution
With the aim of establishing a uniform regulatory framework and legal requirements, the EU is planning Europe-wide legal regulation of all applications that are directly related to the use of artificial intelligence. The project is bundled in the "White Paper on Artificial Intelligence in the EU". Only a few applications are to be classified as unacceptable from the outset. However, all applications that directly influence people or groups negatively by revealing possible weaknesses on the basis of formal data (e.g. in application documents) are affected.
On 21.4.2021, the EU Commission presented a "Proposal for a Regulation on a European Approach to Artificial Intelligence", which all developers and users of AI applications can follow in future to ensure proper handling of this technology.
All in all, the draft law is likely to be a tightrope walk between regulation and overregulation. It therefore remains to be seen how the EU can implement the enormous portfolio of requirements in dealing with the technology.
The proposed Europe-wide regulation must not be misunderstood and equated with an unwarranted intervention in the potential of artificial intelligence.
Regulation is not bad per se; it can help to find the right socially accepted framework of values. It also creates clarity for companies to get off to a good start. Our AI in Control framework helps with this.