AI@HPI 2024

As artificial intelligence (AI) increasingly integrates into various aspects of our lives, addressing the ethical implications of its use becomes more crucial. Algorithmic decisions are often made at such a massive speed and volume that human oversight is no longer practical. In this panel, experts from various disciplines will explore how biased data and a lack of diversity in development can lead to discriminatory outcomes. This is particularly troublesome for recent AI models with many millions of parameters that are far too complex to be interpretable by humans. The discussion will focus on how biases arise in datasets and algorithms, the ethical standards that need to be established, and how to create fair, inclusive, and transparent AI systems. Both technical solutions and societal and business approaches will be examined to ensure fairness, inclusion, and transparency. The aim is to develop practical strategies to minimize discrimination and promote equal opportunities through ethically responsible AI systems. 

In the wake of the COVID-19 pandemic, the need to trust AI in global health has become increasingly important. As we navigate global challenges such as poverty, conflict, and societal polarization, trust in AI and its role in health management is under scrutiny. This panel will explore AI's potential to enhance global public health, improve crisis preparedness, and combat misinformation. Key questions include: Can AI truly support the common good without human oversight? How can we balance technological progress with preserving essential human values? Beyond the health sector, trust in AI is essential for its broader adoption. Health, as a key application area, highlights the need to build and maintain public trust to ensure AI's positive impact. This discussion will focus on the ethical frameworks needed to guarantee AI's beneficial role, while preserving public trust and human dignity in global health initiatives. 

In the digital age, artificial intelligence (AI) is increasingly influencing political processes and decision-making. From targeted election campaigns to the analysis of voter behavior, AI has the potential to reshape democratic systems. While technology offers new opportunities to optimize electoral processes, it also raises significant concerns regarding transparency, fairness, and the risk of manipulation. Some campaigns have been launched by foreign actors seeking to spread disinformation and propaganda. This panel will explore how AI can impact democratic integrity, with a focus on its effects in U.S. elections, state elections in Eastern Germany, and other significant upcoming political events, including the German federal elections. This discussion will address the role of regulation in ensuring fair and transparent elections and measures to safeguard election processes from manipulation and disinformation. Our aim is to develop ethical and legal frameworks that preserve the integrity of democratic processes in the AI era and to ensure that technological advancements contribute positively to democratic governance. 

While the ability to leverage vast quantities of data has been a key driver of recent progress in AI, the intellectual property (IP) rights associated with such training data have often been neglected. This panel explores how IP can be used legally and ethically in AI development, with a particular focus on challenges in the arts and media sectors. Central to the discussion is the legal use of copyrighted materials, such as images, music, films, and texts, in AI models. Experts will discuss how to respect IP rights while also providing comprehensive datasets for AI training. The panel will also examine the specific impacts on the arts and media sectors, especially regarding the rights of artists, authors, and media companies. Strategies for maintaining licensing requirements and protecting creative rights, without hindering innovation, will be explored. The discussion's aim is to develop methods and strategies for fair use of IP in AI training. Participants will gain insights into how the need for data and creative content can be balanced with the protection of IP rights. 

The EU AI Act represents a landmark regulatory framework aimed at ensuring the ethical and responsible development of artificial intelligence (AI) across Europe. This panel will explore the transformative impact of the EU AI Act on workflows, the economy, and global AI regulation standards. Key topics include the Act's provisions on risk classification, transparency, and enforcement mechanisms. Panelists will discuss how these regulations affect various industries — from technology to manufacturing — and their implications for business operations and innovation within Europe. The discussion will also cover the broader scope of sustainable AI regulation, including its role in promoting ethical practices, protecting consumer rights, and maintaining international competitive advantages. A crucial aspect of the discussion will be ensuring that the AI Act avoids repeating past regulatory mistakes and fosters a cohesive understanding and implementation across all EU member states.  

Experts, including policymakers and legal professionals, will share strategies for navigating this evolving regulatory environment while driving innovation and ensuring consistency throughout the EU. Additionally, the panel will examine potential global trends and the future of AI regulation beyond the EU, considering the possibility of international harmonization of AI standards. 

The rapid pace of progress in artificial intelligence (AI) has far-reaching implications for all industries. In this context, there is a need to move quickly to keep up with the competition, while at the same time upholding ethical standards and updating ethical best practices to consider the latest technical developments. This panel will explore how ethical principles, and a sense of responsibility can be integrated into the development and implementation of AI systems. Experts from various industries will discuss how principles such as fairness, transparency, and data protection can be implemented in practice to minimize the risks and potential negative impacts of AI use. A particular focus will be on the development of guidelines and best practices that companies and organizations have adopted to use AI technologies responsibly and sustainably, even in the absence of stringent regulatory frameworks. The discussion will highlight how industry leaders and civil society have sought to proactively establish ethical standards and self-regulatory measures. This includes examining how these practices reflect their commitment to ethical AI use and how they shape the future of AI deployment beyond formal regulations.