AI agent security: PSP at the University of Parma

On 3 December, our Head of AI will be a guest at the University of Parma to give the seminar “AI agent security: autonomy, verification and human oversight”.

The presentation will bring the experience we gain every day at PSP into an academic setting. We develop and integrate AI-based solutions, dealing directly with the opportunities and challenges these technologies introduce into business processes.

Why AI agent security matters

AI agents do more than provide answers: they can carry out tasks, use tools and support software development within defined processes. This autonomy makes it necessary to establish which actions are permitted, how to verify their outcomes and when a person needs to intervene.

Security therefore concerns both the protection of data and working environments and the reliability of the actions performed by these systems. The seminar will address opportunities and risks with a practical focus on transparency, traceability and human oversight.

Autonomy, verification and human oversight: seminar topics

The seminar will explore some of the most relevant topics in artificial intelligence today, including agent autonomy, data protection, oversight of actions performed by AI systems, and ways for organisations to introduce these technologies responsibly and in ways that can be verified.

Drawing on real cases and practical experience, the seminar will examine risks, limitations and good practices for integrating AI agents into development processes and business activities while maintaining transparency, traceability and human oversight.

Bringing PSP’s experience into the university discussion

At PSP, we develop and integrate AI solutions into business processes. This work means dealing directly with the possibilities offered by automation and the conditions needed to introduce it responsibly and in ways that can be verified.

The seminar will provide an opportunity for business expertise and the academic community to come together and share risks, limitations and good practices useful to companies and organisations.

Archive photograph of the University of Parma Mathematics Building: AMs / Wikimedia Commons, public domain.