GDM AI Control Roadmap
Arxiv
pdf
2026-07-01T00:00:00
arXiv Paper — PDF not available.
Only the Executive Summary is available here. To read or download the full paper, visit the
arXiv abstract page.
Abstract
AI agents are rapidly becoming the engine of innovation at frontier AI companies, helping with AI research and development, strengthening cyber-defence, and advancing scientific discoveries. As these agents become more tightly integrated into our systems, we recognise that unlocking their full potential requires rethinking how we do security. We should not assume that AI agents are always perfectly aligned, but should instead build in multiple layers of defence.
Loading executive summary...