Artificial intelligence: This is how the Ethics Council assesses the use of AI

Opinion on 287 pages
“Artificial intelligence must not replace humans” – this is how the German Ethics Council judges the use of AI

The AI ​​ChatGPT causes a sensation


How should humans and learning machines work together in the future? The subject of artificial intelligence is not just the title of the current issue star, and the Ethics Council has also dealt with it intensively. And took a closer look at the use in four specific areas of social life.

It is not often that the topic of the current star-Title is so important that the German Ethics Council is also dealing with it. “What are we creating right now?” he asks star this week and describes the opportunities and risks of artificial intelligence (AI). The opinion of the Ethics Council is a little more extensive. In a 287-page report, the 26 members explore where they consider the use of AI-based systems to be unproblematic and which issues they have concerns about.

The statement was initiated in October 2020 by the then President of the Bundestag Wolfgang Schäuble (CDU) and is entitled “Human and machine – challenges posed by artificial intelligence”.

The central message of the work was announced by the chairwoman of the German Ethics Council, Alena Buyx, at a press conference in Berlin on Monday. “The use of AI must augment human development, not diminish it,” said Buyx. “AI must not replace humans.” These are the guidelines for evaluating the interaction between humans and AI.

The Ethics Council is an independent body in Germany that deals with ethical issues and challenges in the fields of science, medicine and healthcare. The 26 members are appointed by the President of the German Bundestag. The Bundestag or the Federal Government can commission the Ethics Council to advise on certain topics, just as Wolfgang Schäuble did two and a half years ago.

Artificial intelligence: assessed for use in medicine, education, communication and administration

Specifically, over the past two and a half years, the members have taken a closer look at the potential of AI in four areas: medicine, school education, public communication and opinion-forming, and public administration. It has been shown that the assessment of AI must always be context-, application- and person-specific. The delegation of activities to machines could “have very different effects for different groups of people, actors and those affected,” explained the spokeswoman for the working group, Judith Simon. The key question for the ethical assessment is therefore whether the scope of action of the people involved is expanded or reduced through the use of AI.

For the medical sector, for example, the Ethics Council lists reasons why the use of AI could make sense: With the help of AI, supply bottlenecks due to staff shortages could be alleviated and more precise diagnoses could be made. When developing and using AI products, however, a loss of medical competence must be avoided. The privacy of patients must be reconciled with the intensive use of data in medical research. “A complete replacement of the medical specialist by the AI ​​system endangers the patient’s well-being,” says the statement.

According to the recommendations, the use of AI in school education should not be controlled by technological visions, but should be based on fundamental educational ideas. It should also be limited to items that have been shown to enhance learners’ competencies and social interactions.

In the area of ​​public communication and opinion-forming, the Ethics Council recommends, among other things, further development of the rules for online platforms with regard to the selection and moderation of content as well as personalized advertising and data trading. He also calls for better access to platform data for research and recommends considering the establishment of a digital communication infrastructure under public law.

Do not blindly follow machine recommendations

For the use of AI by public administration, citizens would have to be protected from discrimination. Machine recommendations should not be followed blindly. Furthermore, individual case considerations as well as the rights of inspection and objection of those affected would have to be guaranteed.

Wherever you draw the line in the four areas, the ethics council’s central message was summarized by its deputy chairman Julian Nida-Rümelin: AI applications cannot replace human intelligence, responsibility and evaluation.”


source site-5