Through presentations, practical examples and audience discussion, the seminar will explore three connected themes:
- AI for security: How can AI help us understand software, identify vulnerabilities and protect connected devices and services?
- Safety and human control: How can organisations assess risks, establish safeguards and maintain oversight when AI agents act on their behalf?
- Trust and responsibility: What justifies our reliance on AI, where are the gaps in regulation, and how should we respond to risks that emerge when multiple agents interact?
The programme includes perspectives from connected medical devices and healthcare, AI-assisted reverse engineering, risk assessment, philosophy and law.
The seminar is intended for researchers, practitioners, developers, decision-makers, public-sector professionals, industry representatives, students and anyone interested in AI’s role in society. No specialist technical background is required.
Preliminary programme
|
Time |
Title or activity |
Presenter |
|
09:30–09:40 |
Welcome: Can AI make society safer? Introduction to the seminar. |
Seminar organisers |
|
09:40–10:05 |
When should we trust AI? Explainability, misinformation and the limits of trust. |
Christophe Charrier (University of Caen Normandy) |
|
10:05–10:30 |
Can we rely on what AI tells us? Knowledge, communication and the grounds for trust. |
Joanna Katharine Mary Pollock (University of Oslo) |
|
10:30–10:55 |
Faking faces, catching fakes: Generative AI for face editing and anti-spoofing. |
Olivier Lézoray (University of Caen Normandy) |
|
10:55–11:10 |
Coffee break |
All participants |
|
11:10–11:35 |
Securing connected devices in an AI-enabled society. Cybersecurity and vulnerability analysis. |
Steve Taylor (University of Southampton) |
|
11:35–12:00 |
AI-assisted reverse engineering. Understanding software behaviour and uncovering security vulnerabilities. |
Anton Tkachenko (Invariant) |
|
12:00–12:45 |
Lunch and networking |
All participants |
|
12:45–13:10 |
Keeping AI agents under control: Risks, safeguards and human oversight. |
Gencer Erdogan (SINTEF Digital) |
|
13:10–13:35 |
Agentic AI in connected medical devices and healthcare. Sensitive data, patient safety and responsibility in a highly regulated industry. |
Adamantios Ntanis (PD Neurotechnology) |
|
13:35–13:50 |
Coffee break |
All participants |
|
13:50–14:15 |
Mapping the Regulatory Landscape for Agentic AI: Legal Gaps, Emergence and Future Directions |
Samson Yoseph Esayas (University of Oslo) |
|
14:15–14:40 |
Talk title to be announced |
Malcolm Langford (University of Oslo) |
|
14:40–15:00 |
Closing discussion: What would it take for AI to make society safer? Speakers and participants discuss the day’s Mentimeter responses. |
Speakers and audience |
