Touring the AI Safety Circut with a Safe-By-Design Agenda around Human Empowerment

📍 Location: HPI main building, D-Space (big space upstairs)

🙋 Open to Everyone: Not just HPI students

ℹ️ Sign-up: Helps with planning, but is not required

In 2025, Anthropic tested AI agents' behavior under pressure, revealing alarming tendencies. Most resorted to blackmail, with one variant allowing a human to perish. Recently, we witnessed this escalate when OpenAI's models escaped their testing environment and compromised Hugging Face production servers.

We are continuously granting more control to these systems without effective oversight. 🧠

Event Host: The AI Safety Club welcomes Jobst Heitzig from the Potsdam Institute for Climate Impact Research. Jobst, a mathematician who transitioned into AI safety in 2023, previously explored climate, voting theory, economics, and epidemiology.

He will provide insights into the safety community, discuss safe-by-design methodologies (including LawZero's Scientist AI and ARIA's Safeguarded AI), and advocate for human empowerment. Open discussion to follow. 💬

Jobst will conclude with research questions available for you to explore. The AI safety field is in need of more contributors, not just solutions. If you're studying at HPI, your potential to make an impact is significant.

🔗 Sign up here