Safety · Concept
Perspective-Aware Framework for Reproducible ML Evaluation and Alignment
Thesis proposal to stop plurality-vote label aggregation that erases disagreement and minority perspectives in AI evaluation and alignment.
Forecast horizon
Connections
Connections · 2
How this node ties into the rest of the map, and the evidence behind each link.
to · supports 4/10
Plurality of Human Preferences for AI Systems Framework argues evaluation and alignment must preserve minority perspectives instead of collapsing labels by plurality vote.
+4 growth to · supports 4/10
Constitutional Coverage Trilemma in AI Governance Perspective-aware evaluation rejects single consensus labels, aligning with evidence that constitutional value demand is broad and plural.
+4 growthSignal sources
Signal sources
Dated facts from primary sources in this direction.
US evaluation centre Jun 2025
In June 2025 the US AI Safety Institute was renamed the Center for AI Standards and Innovation (CAISI), pivoting toward security, standards and adversary-model assessment.
NIST → Frontier safeguards May 2025
Anthropic activated its ASL-3 deployment and security standard with Claude Opus 4 on 22 May 2025 — the first real-world trigger of a responsible-scaling tier, focused on blocking bio-weapon uplift.
Anthropic → Cross-border testing 2025
The International Network of AI Safety Institutes (launched Nov 2024) ran a third joint testing exercise focused on agentic AI systems across cyber and fraud strands.
European Commission — AI Office →