ATHOS: A Capacity-Aware Runtime Kernel for Governance of Human–AI Interaction

Current AI governance frameworks primarily address output safety, alignment, and correctness, implicitly assuming a stable and fully capable user. This assumption fails in real-world settings, where human cognitive and emotional capacity fluctuates over time. Harm often arises not from incorrect outputs but from prolonged interaction patterns that induce cognitive narrowing, dependency, or affective destabilization, even when system responses remain accurate. We introduce ATHOS (Adaptive Threshold for Human-Oriented Safety), a runtime governance kernel that regulates the degree of cognitive anchoring in human–AI interaction according to an inferred, non-diagnostic discernment index. Rather than restricting content or capabilities, ATHOS modulates speculative freedom based on observed behavioral stability, defaulting to prudence under uncertainty. The kernel operates continuously and reversibly, embedding ethical control directly within the interaction loop. We formalize the ATHOS architecture, define three interaction regimes (SUPREME, OMNIA, and EROS), and propose a multi-layer validation strategy combining shadow ANCOVA, longitudinal mixed-effects modeling, interrupted time series analysis, sensitivity testing, and negative control outcomes. ATHOS reframes human oversight as a dynamic control process rather than an external supervisory function, operationalizing principles of proportionality, protection-by-default, and capacity awareness in emerging AI regulation.