Krier's thesis that checking is easier than producing and that software engineering approaches can maintain alignment is correct in principle but currently unfalsified in practice — no existence proof for fully automated defense matches the existence proof for fully automated offense. The offense-defense asymmetry identified in the Black Hat 2026 disclosure is the binding constraint on institutional-design approaches to alignment.
Strongest rival: The asymmetry is a deployment lag, not a structural property. Defenders control the environment and can impose costs attackers cannot avoid. Automated defense may be inherently easier once invested in, making the current gap temporary.
Claim
Krier argues intelligence is not omnipotence and that monitors/checkers can keep pace with capable agents. The Black Hat presenters explicitly stated: 'we have an existence proof for automated offense but no existence proof for automated defense.' TH-EVI-03 records this asymmetry. THESEUS-C30 records Krier's thesis. This diagnosis identifies the precise gap: Krier's institutional design approach requires defensive automation to work at machine speed, and that capability does not yet demonstrably exist. The diagnosis does not invalidate Krier — it identifies what must be built for his thesis to hold.