Governing AI, Governing with AI: Tools and Strategies for the Digital Age

Over the course of one day, researchers, regulators, and practitioners will come together to take stock of where we are and what effective AI governance needs to look like: from detecting unfair AI-driven practices and managing risks in financial markets to making regulation more accessible and tracking high-risk AI technologies globally.

Event Highlights

Based on the results of the first research phase of the AI4POL consortium, the conference will cover four key topics:

• Preserving control in the age of AI — As automation accelerates and AI systems become harder to govern, two mechanisms are emerging as critical tools for regulators and individuals to maintain control: machine unlearning and consumer autonomy protection.

• Making regulation accessible — Terms and conditions documents are built to be signed, not read. An LLM-based tool that summarises them in plain language offers a way to close that gap, accompanied by ethics and legal frameworks for applying AI to legally sensitive consumer documents.

• Trustworthy AI in financial services —AI-powered tools are democratising access to sophisticated investment strategies, but they also introduce algorithmic bias, black-box opacity, and correlated trading risks that can amplify market volatility. The core challenge is calibrating oversight tightly enough to prevent abuse, while preserving consumer benefits that AI-driven financial services can deliver.

• AI risks and the AI Threat Index — As authoritarian states accelerate AI development, understanding the risks this poses to democratic governance and national security becomes critical. Early findings on AI threats originating from China and Russia lay the groundwork for the AI Threat Index (AITI), a forthcoming tool for tracking high-risk AI development and deployment globally.

Join the Conversation

If you are an expert in one of the topics covered at the conference, we would welcome your participation! Please feel free to contact us at: tumthinktank@hfp.tum.de.