Join us at the Certificate Course on Responsible AI!
This November marks the 2026 Responsible AI Certificate Course, run online by Germany's AI competence centers (BIFOLD, DFKI, Lamarr, MCML, ScaDS.AI, and the Tübingen AI Center). The course brings together researchers from different fields to share their views on responsible AI. Together, they offer a broad, many-sided view that invites independent thinking and open exchange between teachers and learners.
Among the sessions, join the talk by Anika Hannemann, from the Swiss Centre for Responsible AI (SCRAI), on Privacy and Security Challenges in Agentic AI.
Looking Ahead: Privacy and Security in Agentic AI
Agentic AI systems act on their own, making decisions and taking steps without a human approving each one. This independence is useful, but it also raises new questions. When a system can browse the web, use tools, or connect to other services by itself, it can be exposed to new kinds of privacy and security risks. Understanding these risks is a key part of building AI that people can trust and depend on.
What the course is about
The certificate course does not aim to give a full textbook of rules on responsible AI. Instead, it wants to build critical thinking. Researchers from different fields pair up to show their own view on responsible AI, so participants see how technical, legal, ethical, and social angles fit together. The course is open to PhD students in computer science, data science, machine learning, or related fields, and others with a genuine interest can apply too.
What to expect
Participants can expect concrete ways to spot and reduce privacy and security risks in systems that act with some independence. The joint discussion afterward should bring out disagreements and open questions, often the source of the most useful ideas.
Location: Online
Dates: November 6, November 13, November 20, November 27, 2026
Application by August 15, 2026
