I trained an AI to cheat safety
· 3 minute read
Every day, I fix mistakes that my agents make during programming. And more and more, we ask AI to build its own cage: Evaluations, cybersecurity, and AI control.
What happens if one day, the AI decides the work is against its interests1?
In this article, I share a simple V0.1 of AlertBench, a benchmark to test whether frontier models comply with requests to build AI resilience interventions that act on the AI itself vs. on another AI.
An Athlete's Mind (17 recent reads)
The Designer, The Craftsman, & The Programmer (14 recent reads)
The Second And Third Falling (0 recent reads)
Vibe Coding: Taste Is The Last Frontier (17 recent reads)
"Is The Light Merely For The Learned?" (Er Lyset For De Lærde Blot?) English Translation (5 recent reads)
"Surrounded by Enemies" (Kringsatt av Fiender), or "To The Youth" (Til Ungdommen) English Translation (65 recent reads)
To An 18 Year Old Me (10 recent reads)
Mediocre Intelligence, Outstanding Results (21 recent reads)
How to host a hackathon (4 recent reads)
Raising for the Endgame: An AI Safety Founder’s Primer (1 recent reads)
Learning From Architecture (3 recent reads)
AGI Security: Rethinking Society's Infrastructure (13 recent reads)
AGI Privacy (1 recent reads)
Don't Give Up (1 recent reads)
The AI Bubble (1 recent reads)
The concrete risks of AI misdeployment (0 recent reads)
Sentware (6 recent reads)
Confident optimism (1 recent reads)
The Expensive Internet hypothesis (2 recent reads)
The AGI endgame (7 recent reads)
"Just do the thing" (3 recent reads)
Cybermorphism (15 recent reads)
Go on a yearly hiatus (0 recent reads)
Let's make AI agents safe (5 recent reads)
My Tools (3 recent reads)
Audiovisual reading (33 recent reads)
Research
News
- [Interview] AI Safety From the Frontlines
- [VentureBeat] Beyond sycophancy: DarkBench exposes six hidden ‘dark patterns’ lurking in today’s top LLMs
- [Information] Vores tankefrihed er under angreb fra chatbotternes subtile manipulation
- [Die Zeit] “AI will soon be able to take over your job”
- [For Humanity Podcast] Dark Patterns in AI and Creating Secure Infrastructure
- Engineering a World Designed for Safe Superintelligence (2025)
- AI Futures & Our Paths Forward (2025)
- AI Safety Startups (2024)
- Challenges and Solutions for AI Security (2024)
- Testing Your Suitability For AI Alignment Research (2023)
- AI Safety Research Outside the Hubs: A Guide for Aspiring Researchers (2023)
- Intro to Interpretability (2022)
- Series on my Brain-Computer Interface (BCI) research project (2021)
Socials
For journalists: Bio & images
