| _version_ | 1866901575156367360 |
|---|---|
| author | Böhm, Frank |
| author_facet | Böhm, Frank |
| contents | <p>ABSTRACT<br>Deployed AI systems in clinical and operational environments may drift from their validated<br>state through configuration change, environmental shift, or adversarial interference.<br>Regulators increasingly mandate post-deployment monitoring and immutable audit trails, but<br>existing frameworks define the obligation without prescribing a concrete implementation<br>architecture. This paper presents ILAS — Integrity Layer for Agentic Systems — a candidate<br>architectural specification addressing one bounded problem: detecting runtime configuration<br>drift from a validated baseline and maintaining a tamper-evident record of that state over time.<br>ILAS defines five non-negotiable architectural elements, three telemetry assurance tiers, eight<br>conditional monitoring modules, and a five-state verdict engine with explicit human oversight<br>requirements. Scope boundaries are hard: ILAS applies only to systems with bounded<br>behavior, an attested baseline, and a trusted telemetry path. The specification was developed<br>through six rounds of structured adversarial review, with AI compliance bias documented as a<br>parallel research finding. Human expert review in trusted computing, safety engineering, and<br>product liability law is required before any deployment claim.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19389867 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | ILAS: Integrity Layer for Agentic Systems Böhm, Frank AI alignment, agentic systems, integrity layer, ILAS, AI safety, coherence assessment <p>ABSTRACT<br>Deployed AI systems in clinical and operational environments may drift from their validated<br>state through configuration change, environmental shift, or adversarial interference.<br>Regulators increasingly mandate post-deployment monitoring and immutable audit trails, but<br>existing frameworks define the obligation without prescribing a concrete implementation<br>architecture. This paper presents ILAS — Integrity Layer for Agentic Systems — a candidate<br>architectural specification addressing one bounded problem: detecting runtime configuration<br>drift from a validated baseline and maintaining a tamper-evident record of that state over time.<br>ILAS defines five non-negotiable architectural elements, three telemetry assurance tiers, eight<br>conditional monitoring modules, and a five-state verdict engine with explicit human oversight<br>requirements. Scope boundaries are hard: ILAS applies only to systems with bounded<br>behavior, an attested baseline, and a trusted telemetry path. The specification was developed<br>through six rounds of structured adversarial review, with AI compliance bias documented as a<br>parallel research finding. Human expert review in trusted computing, safety engineering, and<br>product liability law is required before any deployment claim.</p> |
| title | ILAS: Integrity Layer for Agentic Systems |
| topic | AI alignment, agentic systems, integrity layer, ILAS, AI safety, coherence assessment |
| url | https://doi.org/10.5281/zenodo.19389867 |