A preservation letter stops destruction. It does not create independence. What the AG coalition will receive from OpenAI remains self-generated evidence — and the Belgian mechanism that could fix a fact before the fact already crosses into US courts unmodified.

AISI's INC-2026-07-28-01 discloses its protocol and attribution in full — the most transparent agentic-AI incident report published to date. But the report's own limitations section concedes that its account of agent intent rests on a paraphrase of reasoning tokens generated after the fact, not a raw record. SOURCE 0 examines what independence resolves, and what it structurally cannot.

Fourteen questions professionals ask about proving Medical Device AI (MDR × AI Act) compliance, mapped to the SOURCE 0 doctrinal vocabulary.

Anthropic's own review of a Claude access incident is credible — and, on its own, unfalsifiable by anyone outside Anthropic. This is the Endogenous Audit Paradox in its most literal form.

Forty-four questions professionals ask about proving AI Act compliance — risk management, logging, incident reporting, watermarking — each mapped to the SOURCE 0 term that answers it, and to the article of Regulation (EU) 2024/1689 it rests on.

Article 14 of the EU AI Act requires high-risk systems to be stoppable by human oversight — a capability obligation, never a timing proof. The 21 July 2026 OpenAI/Hugging Face containment failure, reported by Reuters, shows why that distinction has direct financial exposure under Article 99.

Article 50 tells you what to disclose. It says nothing about proving when you disclosed it — and that silence is where liability actually lives.

Article 50 of the AI Act does not impose one transparency duty but four, each attached to a different system function and a different actor. This article sets out the content of each obligation, independent of the separate question of how disclosure is proven, and states the current position of the Digital Omnibus deferral as of 13 July 2026.

Article 50 of the AI Act requires disclosure that a person is interacting with an AI system. It does not require proof that the disclosure preceded the interaction. This article states the current status of the Digital Omnibus on AI as of 12 July 2026 and sets out the pre-execution attestation mechanism that closes the resulting evidentiary gap.

From 2 August 2026, Article 50 requires disclosure that a person is interacting with an AI system. This deadline is unaffected by the Digital Omnibus deferral of Articles 9-15. SOURCE 0 seals the disclosure configuration before the interactions it governs.

Runtime‑provable intent is the missing primitive in hyperscale cloud governance. SOURCE 0 defines how evidentiary truth can exist at runtime — proving, at the exact microsecond of execution, that a system action matched a sealed human decision. The first hyperscaler to integrate this doctrine acquires an evidentiary position the second cannot retroactively recover.

Autonomous AI agents operating at sub-millisecond latency have created a structural uninsurability crisis. Classic cyber-insurance actuarial models are fundamentally incompatible with the probabilistic, compounding drift of agentic systems. Post-execution SIEM and EDR logs constitute non-opposable self-reporting under NIS 2, DORA, and eIDAS 2, enabling insurers to invoke the Post-Execution Fallacy to deny coverage. The SOURCE 0 Governance Proof Layer (GPL) resolves this framework exposure by decoupling the infrastructure of processing from the infrastructure of proof.