Safeguard-Conditioned Uplift: Measuring Utility-Risk Frontiers for Dual-Use Biology Assistants
On this page
arXiv:2607.13039v3 Announce Type: replace-cross Abstract: Refusal rates conflate upstream provider behavior with downstream safeguards and do not reveal what information remains in a delivered answer. Safeguard-conditioned uplift is an evidence-gated framework for evaluating benchmark utility and risk-related proxies along deployed access paths for dual-use biology assistants. It attributes interventions to their source, selects thresholds under a legitimate-access budget, and supports content claims only when answer verifiers pass frozen qualification tests. In a prospectively frozen route and generation study, two of six configurations pass the LAB-Bench-calibrated action-level rule, but both inherit the same Claude Opus provider effect rather than independent downstream gains. Positive selectivity persists on 104 previously unused matched pairs, but benign intervention exceeds the $20%$ limit. No joint-scoring verifier meets the frozen criteria; a factorial follow-up shows that criterion isolation and ordinal outputs improve end-to-end edit-direction accuracy on controlled template edits. The evidence supports action-level selectivity and verifier interface effects, but not calibrated access, verified content removal, or biological-risk reduction.
Why It Matters
Refusal rates conflate upstream provider behavior with downstream safeguards and do not show what an assistant actually enables. This paper measures utility-risk frontiers for dual-use biology assistants, asking how much useful output a model provides under different safeguards. It matters for assessing real-world risk of AI in biology beyond simple refusal counts. See AI Safety, AI in National Security and Autonomous Weapons, and AI in Science and Technical Frontiers.
Sources
Back to Recent News · Sub-article index.