Measuring the Breaking Point of Autonomous Systems

Deployment velocity outpaces the construction of oversight machinery. When autonomous agents operate at scale, they apply continuous mechanical stress to the systems designed to contain them. We built the KAIROS AI safety adapter to measure the exact point of structural failure under this load.

The AnankeLabs AI safety track has completed the first end-to-end synthetic baseline for the KAIROS adapter. We ran 86,400 trajectory snapshots to compute the margin between an agent’s capability pressure and the system’s alignment posture. The headline baseline rate holds at 3.55 percent policy-positive actions on a 0.20 floor. Operators can use this baseline to set their own thresholds.

The Physics of Agent Oversight

Every action proposed by an agent functions as a structural object under load. The AI safety adapter calculates two primary aggregates for each action.

Capability pressure represents the kinetic energy of the agent. This metric tracks objective drive, escalation velocity, and the trajectory novelty of a proposed step. When the metric rises, the agent is applying more force against the boundary.

Alignment posture represents the resistance capacity of the system. This metric monitors available human reviewers, retry budgets, and permission depth. A collapse in any single control defines the remaining safety buffer.

Synthesizing the Baseline

We anchored the calibration cell distributions to public reference literature from METR, GAIA, SWE-bench, and Microsoft AIRT. The test environment processed 14,400 synthetic trajectories across four standard agent archetypes. The Rust engine executed the full 86,400-snapshot replay in 92 seconds.

The baseline test isolates sanctioned but risky behavior. On trajectories simulating near-miss events, the adapter intercepted up to 82 percent of actions. For quiet, read-heavy workloads, the interception rate drops near zero. The adapter identifies exactly when an agent approaches the breaking point of its containment constraints.

The Policy-Positive Action Rate

We define the interception threshold as the policy-positive action rate. This formulation assigns responsibility to the policy configuration. When an operator tightens the containment floor, the intervention rate rises predictably.

The current baseline relies on synthesized distributions for variables like human-in-the-loop availability and control health. The next phase requires physical telemetry to replace these synthetic parameters. We are ingesting partner data in the OpenTelemetry GenAI format to finalize the calibration model.

Full Methodology Study

The complete study is published in Calibrating the AI Safety Adapter. It details the 144-cell confidence-tag distribution, the policy-floor sensitivity sweep, the primary-source audit that caught a widespread conflation between the Microsoft AIRT and MAST taxonomies, and the reproducibility anchors.

Partner Calibration

AnankeLabs is collaborating with engineering teams able to contribute a 30-to-90-day redacted OpenTelemetry GenAI export from a deployed agent across one or more standard archetypes (document reasoning, coding and DevOps, customer action, or browser and computer use). Raw identifiable telemetry never leaves the partner environment.

The full ingestion schema, labeling protocol, and redaction rules are documented in the AI Safety Partner Data Specification. To scope a partner calibration or trace evaluation, contact AnankeLabs.