MeMo:KI – Opinion Monitor Artificial Intelligence

These and many other questions have been asked frequently in recent years.
Findings on attitudes toward AI are by no means uniform: According to Bitkom (2021), there is a rather positive basic attitude toward AI technologies among the population. The figures indicate that AI is perceived exclusively or mostly as an opportunity by almost three quarters of those surveyed. Only 13 percent of Germans perceive AI as a threat. The Eurobarometer (2020) shows that for around half of those surveyed (51%), measures are required on the part of policymakers to ensure an ethical approach to AI. A large majority (80%) of respondents want to be informed about whether a digital service or mobile application uses AI in different situations. Elsewhere, however, there is a greater emphasis on risk assessments. YouGov points to a more negative perception of AI in Western countries. Here, risk perceptions differ between specific application fields (YouGov, 2021). Recently, the Bertelsmann Foundation reported a large knowledge gap in Germany, according to which large parts of the population assume that they have only limited knowledge about algorithms and AI, and that they can hardly assess the consequences of their use (Bertelsmann Foundation, 2022).

In order to shape technology in a way that is compatible with democracy, however, decision-makers should be aware of the ideas of the population and (civil) society. Why? There are two answers to this: First, people in decision-making positions tend to strengthen their own position by drawing on a personal impression of “public opinion”. This can have far-reaching consequences for the use and design of AI, for example when public money is spent on the promotion of AI research or further education for citizens on the basis of a feeling. Our project MeMo:KI provides a reliable database as a counter-design to anecdotal knowledge.

In addition to the population survey, this also includes an analysis of media coverage on the topic of AI. For the 32 print and online media with the highest coverage, monthly reports are made on which top topics are most strongly included in media coverage. In the past, it has been shown that the thematic focus in media coverage is primarily on economic efficiency gains through AI. Reports on opportunities for economic progress predominate, while risks are mentioned less (Fischer & Puschmann, 2021).

In various papers from country to EU level, institutions also claim human-centered AI development that complies with ethical principles and label this as a unique selling point of AI “made in Europe”. However, such a framework is not a foregone conclusion. Democratic and social compatibility is not ensured by political institutions writing it into position papers. If policymakers take their own demands seriously, they depend on the support of an active and enlightened (civil) society that can contribute ideas to the shaping of digitization policy and, above all, demand their implementation. Otherwise, the development as well as implementation of AI technologies will remain the preserve of interested industry, whose primary goal is to maximize profits and increase efficiency. In a long-term study of public and published opinion, MeMo:KI therefore investigates the extent to which the conditions for opinion formation on the topic of AI in the population are given and how these conditions change over time.