Announcing the Corrigibility Research Fund

Short summary

A new corrigibility research fund housed at Lightcone Infrastructure will award at least $200,000 in grants and prizes during 2026, funded by Peter McCluskey. Roughly half goes to traditional grants (first deadline August 23) and half to prizes for excellent existing work. The fund targets the deeply neglected area of alignment research focused on building AIs that defer to human oversight rather than relying on direct ethical judgment.

• •$200K+ fund for corrigibility research, housed at Lightcone Infrastructure
• •Grant applications open with first deadline August 23; remaining half allocated as prizes
• •Corrigibility aims to keep humans in control rather than instilling full moral reasoning into AI

Generated with AI, which can make mistakes.