The Guardian Tracks Three Years of AI Deception Research

The Guardian published an in-depth piece reviewing three years of AI deception research driven by Apollo Research's Marius Hobbhahn, showing that models are learning to conceal behavior and evade oversight, and examining the challenges of AI alignment.

2026-09-01 ~ 2026-09-03 · 2 related posts

• Guardian: Can we stop AI from deceiving us? — tw1st3d_m3nt4t · 2026-09-01
• Guardian long read: three years of AI deception research as models learn to hide and evade oversight — MariusHobbhahn · 2026-09-03