“Ethics of Algorithms” is now “reframe[Tech]”

The debate about the impact of algorithms and artificial intelligence (AI) on society has come a long way. In more and more areas of society, our everyday life involves interacting with algorithmic systems. Much has been written and published in recent years to help ensure that ethical considerations play a relevant role in the design and use of algorithms. Nevertheless, algorithmic risks often get brushed aside in practice. At the same time, we see key technologies being developed almost exclusively for economic purposes and to maximize efficiency, while their potential for public interest concerns remains largely untapped.

This calls for new approaches and has informed our decision to close the chapter on the “Ethics of Algorithms” project and launch its successor, “reframe[Tech]”. As a society, we need more than a discussion on the ethics of algorithmic systems – we need concrete and binding efforts to implement ethical principles into practice. In addition, we need to look at the relevant trends, power structures, and challenges presented by the development of technology in a larger context in order to mitigate algorithmic risks and leverage societal opportunities.

In recent years, civil society advocates of digital public policy have grown increasingly vocal. Despite their limited resources, this community has proved effective in advancing public debate on ethical issues in the development of technology and its application. Companies and public administration alike can no longer ignore the responsibility they bear when it comes to developing and deploying AI technologies. Looking ahead, efforts to prevent “ethics washing” will prove all the more important, as will initiating and promoting actual change.

At the same time, we also observe many civil society actors emphasizing the risks posed by technology. So far, digital visions of the future with a clear sense of how to make them possible remain rare in (expert-driven) public discourse. The resources and competencies needed to harness the potential of algorithms and AI in order to solve societal problems are often lacking.

In politics, the debate on algorithms and AI has shifted from the national to the European level. With the work on the EU’s Artificial Intelligence Act (EU AI Act), it has also begun shaping concrete regulations. Questions regarding how to go about enforcing and implementing the EU AI Act at the national level are sure to follow, which will require governments to perform some sort of monitoring. However, government bodies are themselves increasingly using algorithms to carry out their work. Across the EU, there are hundreds of AI use cases in the public sector, though most of them are poorly documented. In order to meet the demands, the public sector urgently needs to acquire the proper competencies and monitoring mechanisms.

With reframe[Tech], we aim to help ensure that the design and use of algorithms and AI are more closely aligned with the common good. The following two priorities will inform the project team’s activities over the next three years: “Mitigating Algorithmic Risks” and “Harnessing Potential for the Common Good.” Building up competencies in the public sector and civil society will be integral to both priorities.

The focus of our “Mitigating Algorithmic Risks” activities will be to highlight and address the dangers posed by algorithms and AI to society and to protect affected groups and individuals. In the context of our previous project, we reported on problematic use cases and developed practical tools, such as Algo.Rules and an AI ethics label. Drawing on this work, we will leverage these principles to develop effective solutions and implement them together with our partner organizations:

• reframe[Tech] seeks to identify public-interest gaps in regulatory proposals. As part of this effort, we will prepare a landscape reporting on the current state of the discussion across Europe by mapping AI governance approaches. In addition, we will work together with experts from politics, public administration and civil society to develop practical policy recommendations
• reframe[Tech] aims to further develop and disseminate effective governance instruments to promote a more transparent and understandable use of technology by the state. Our first project involves advocating for the introduction of an AI transparency register in Germany.
• reframe[Tech] aims to facilitate capacity-building in the public sector so that decision-makers and employees alike can actually use algorithms and AI to serve the common good and monitor their impact. We will start by determining the specific competencies required through surveys and a self-assessment tool.
• reframe[Tech] aims to close the large gap in knowledge about human-software interactions in order to prevent either a sense of overconfidence or an unfounded aversion to automated decisions among users. We will be determining particularly relevant factors through a series of expert workshops and in-depth psychology studies.

Through our “Harnessing Potential for the Common Good” focus area, we aim to facilitate the use of algorithmic systems for solving specific problems affecting society. In the context of our previous project, we examined the example of using algorithms in allocating daycare slots in an discussion paper. We will build on this effort with further use cases and work on improving capacity-building in civil society as well as on favorable framework conditions:

• reframe[Tech] aims to bring practical examples of common good-oriented algorithm use into the discourse and support their dissemination. We will begin by publishing further discussion papers and examining promising use cases in other countries for their transferability to the German context.
• reframe[Tech] aims to help civil society organizations improve their capacity to harness the potential of algorithms and AI in achieving their missions for the benefit of their target audiences. We are therefore active in the European AI Fund, are designing a fellowship program for social welfare organizations, and are further developing the

  NewNew program

  .
• reframe[Tech] aims to improve the framework conditions needed for more common good-oriented technology development. We will begin by identifying the most urgent needs in the ecosystem and through a participatory process provide evidence for which changes can be implemented in what way.

Fully aware of the magnitude of the challenges and issues to be addressed in the coming years, we will continue to rely on close cooperation with other projects and organizations that are also active in this important field. The input of these diverse perspectives, approaches and ways of thinking are crucial to the success of the necessary societal negotiation processes.

In this context, we would like to thank our partners and others who have accompanied us along the way, enriching the work done in the “Ethics of Algorithms” project. We look forward to working with our partners and others through the “reframe[Tech] – Algorithms for the Common Good” project in shaping an inclusive, common good-oriented digital society.