Browse by subject:
Paper #1015
DORA, the EU AI Act, and the self-application problem
The paper applies Decision Accounting to the Digital Operational Resilience Act (DORA) and the EU AI Act, showing how a seventeen-field decision record can satisfy documentation obligations under both regulations. It resolves the self-application problem: a tool that scores Decision Accounting records is not high-risk unless it affects decisions about natural persons. The framework reduces documentation fragmentation but requires naming the flawed game, recording the Missing System Theory, and preserving human authority.
OPEN HTML DECK ↗Deck mode is an on-site reading view, not a PowerPoint download.
Theorem status: evidence-traced claim under the cited paper's assumptionsFalsification: show the same game preserving system welfare without changing the payoff structure
KEY FINDINGS
THEOREM
Theorem 1 (Missing System Theory): Let N={1,...,n} be agents, P_ij(d) recorded bilateral payoffs for decision d, X(d) system-state variables. If at least one component of X(d) is not recoverable from P(d), then system welfare W(d) is not a function of P(d) alone.
PLAIN ENGLISH
If a decision record only captures bilateral payoffs (e.g., contract terms) but misses system-level effects (e.g., concentration risk, resilience), then you cannot determine whether the decision improved overall welfare.
EVIDENCE & LIMITATIONS
- Theorem status: evidence-traced claim under the cited paper's assumptions
- Falsification: show the same game preserving system welfare without changing the payoff structure
REFERENCES / CITATION STATUS
References section
Detected
Bibliography entries
128
In-text citations
0
Unique citations
118
Footnote markers
118
Citation year span
1689-2025
Source hash
21037b1a55a9
This page reports reference counts measured directly from the manuscript. Full reference entries render only when a curated source chapter carries a public References, Bibliography, Source Notes, Supplemental Reference Archive, Footnotes, or Source-Grounding Ledger section. The site does not synthesize citation entries. Literature-claim verification status: not evaluated.
Loading curated reference section…
EXECUTIVE SUMMARY
The Digital Operational Resilience Act (DORA) and the EU AI Act create a shared problem for financial institutions: regulators require reconstructable decision records, but fragmented documentation often leads to the Missing System Theory. Decision Accounting supplies a seventeen-field record that connects the Missing System Theory to audit practice. The paper formally maps each Decision Accounting field to DORA Articles 5, 6, 8, 11, 13, 17, 18, 19, 28, and 30, and to EU AI Act Articles 9, 11, 13, 14, and 26. It proves that a complete record passing the five-minute/stranger test satisfies the core documentation function of both regulations. The paper then resolves the self-application problem: a tool that scores Decision Accounting record quality is not high-risk merely because it audits high-risk AI documentation. It becomes high-risk only when its intended purpose or actual use places it inside Article 6 and Annex III, such as employment, credit, or essential services decisions about natural persons. The policy implication is that regulated firms should treat Decision Accounting as a control architecture, not a legal safe harbor. The framework reduces documentation fragmentation, but only if the record names the flawed game, records the Missing System Theory, states revenue rather than profit as Π, and preserves human authority over person-affecting uses.
METHODOLOGY
The paper uses formal theorem-proving (Missing System Theory), propositional mapping (field-by-field regulatory alignment), and conditional analysis (self-application problem). It applies the SAPM canon's Decision Accounting framework to two EU regulations, with worked examples and falsification conditions.
SOURCE QUESTIONS
SSRN not yet postedDeck ↗
WHY THIS MATTERS
For the economist
A structural claim about when bilateral optimization degrades the shared system. Read the formal statement and its stated axioms.
For the regulator
The constraint is physical or biological, so disclosure alone will not internalize it. The policy lever is to bound exposure, not to price it away.
For the executive
This is where a privately efficient decision can degrade the system the business depends on. The governance question is which decision records would make that system cost visible before it is normalized.
For the teacher
An on-site HTML deck and the expanded curriculum cover the argument, the evidence, and the measurement. Use the deck as a self-contained class session, then route deeper through the 45-50h core course or 100+h full curriculum.
For the affected community
In plain terms: who gains from the current arrangement, who pays for it, and what rule change would alter that split. The summary states each without jargon.
© 2026 Erik Postnieks · Independent Researcher · Salt Lake City