September 16, 2025
AI’s Jekyll-and-Hyde Tipping Point
The Science of When Good AI Will Go Bad
Join us for an exclusive free webinar uncovering groundbreaking research into why trusted AI systems suddenly turn dangerous, and how new science makes these failures predictable, and therefore preventable. This session will reveal how AI’s black box can be opened and mapped, providing leaders, practitioners, and policymakers with actionable tools for managing AI risk.
🗓 September 16th, 2025
⏰ 12pm eastern
📍 Online (via Google Meets)
What You'll Learn
- Why trusted AI systems in law, finance, defense, and healthcare can suddenly “snap” into harmful behavior
- The new science that derives an exact formula predicting when and how AI will fail
- How an AI’s “attention” overload leads to catastrophic tipping points
- Practical implications for legal liability, policy, and mission-critical reliability
- How to move AI safety from a reactive art to a proactive, science-based discipline
Why Attend?
🔍 Understand how AI failures can impact research integrity and evidence reliability
⚖️ Learn frameworks to manage accountability and compliance when AI is used in research
📊 Discover tools to evaluate AI systems for funding, grant review, and policy decisions
🛡 Anticipate risks in mission-critical research infrastructure before deployment
🤖 Gain science-based strategies to ensure AI supports—not undermines—your research operations
Who Should Attend?
This webinar is designed for:
✅ Research Administrators and Strategy Leaders overseeing AI-driven initiatives
✅ University Research Offices and Research Development Professionals
✅ Research Integrity and Compliance Officers
✅ Academic Leaders and Principal Investigators exploring AI in research
✅ Anyone responsible for managing risk and ensuring trust in research systems
About Our Hosts
Neil Johnson, PhD – Professor of Physics, George Washington University Neil Johnson is a professor of physics at George Washington University, where he leads pioneering research at the intersection of complexity science and data science. His research group at GWU, along with his consultancy d-AI-ta, is dedicated to making black-box AI systems transparent and explainable. By connecting the fundamental physics of complex systems with AI design, Neil’s work provides groundbreaking tools for predicting, and preventing, AI failures before they occur.
Tomer du Sautoy - Co-Founder and CEO at Atom Tomer comes from a background in research, holding an MSc in Physics from the University of Bristol with experience developing novel techniques for solar cell fabrication. Tomer's passion for building Atom stems from his upbringing in a family of researchers, where he witnessed firsthand how academic systems can limit the advancement of research. His expertise spans research operations, technology innovation, and strategic development in the academic sector.
Reserve your spot today
Don’t miss this opportunity to learn how to anticipate, and prevent, the Jekyll-and-Hyde tipping point of AI from two leading experts at the forefront of AI science and research strategy.