Before a decision can be judged, its conditions must be understood.
Research into how AI-assisted decision environments take shape, what meaningful human judgment can still accomplish, and what institutions can reconstruct from the evidence that remains.
An independent
research record.
A decision is more than its recorded outcome. Its formation, the conditions of judgment, and the limits of later reconstruction each deserve examination.
ID∆AC™ Research brings together original essays, conceptual and methodological working papers, and invited independent perspectives. Publications are attributed and versioned; proposed constructs remain distinguishable from empirical findings and operational services.
Intellectual independence
Claims remain open to criticism, alternative models and documented counterexamples.
Traceable publication
Author, document class, edition and durable research record are kept visible.
Research in motion
Context may evolve without silently rewriting a deposited canonical publication.
When the Decision Has Already Begun
AI and the Formation of Decidability
Before a human reviewer receives a recommendation, the situation may already have been filtered, framed and made actionable. This essay asks how a decision field comes into existence—and what that means for oversight, accountability and the limits of formal authority.
Publications & working papers.
Browse published research by subject, document type and date. The catalogue is designed to accommodate further ID∆AC™ work and guest contributions without flattening their authorship or research status.
Showing 5 selected publications · Featured essay shown separately
Preservation of Effective Judgment
A falsifiable research agenda separating opportunities for examination, demonstrated reconsideration and material intervention in AI-mediated decisions.
Conocer el algoritmo no es reconstruir la decisión
An inquiry into the limits of algorithmic knowledge as evidence for reconstructing consequential decisions.
La separación de poderes tras la IA
A research contribution addressing institutional authority and the distribution of powers under AI-mediated governance.
Antes del próximo incidente: identidad del agente, autoridad limitada y continuidad epistémica en sistemas de IA consecuentes
An examination of agent identity, bounded authority and epistemic continuity in consequential AI systems.
The Organisation Has Not Changed. The Decision Has.
How AI may reshape the informational conditions of judgment while formal authority remains unchanged. The essay distinguishes formal authority from decision-shaping capacity.
Questions that connect the work.
How does a decision become possible?
Problem formulation, filtering, ranking and the construction of what becomes visible or actionable before human review.
Read the formative essayWhen can people meaningfully intervene?
Available evidence, demonstrated examination, reconsideration, authority, timing and material action, tested against established comparators.
Explore PEJ working paperWhat can the surviving record establish?
Decision-state reconstruction, attribution, uncertainty and the difference between proof of approval and evidence of effective review.
Read a related public articleIndependent voices. A shared research question.
ID∆AC™ Research intends to host invited essays and external perspectives on a periodic, approximately monthly editorial basis. Contributions may extend, challenge or qualify the research. Publication remains subject to editorial review, accurate attribution and explicit evidential boundaries.
Propose a guest contributionGuest authors retain their own intellectual position. Appearance on the site does not imply endorsement of ID∆AC™ methods, participation in its projects, or agreement with every published argument. No guest edition is announced until confirmed.
Good research leaves room for disagreement.
Found a relevant study, a boundary condition, a competing interpretation or a case that challenges a claim? Send a precise response. We welcome criticism that improves what can—and cannot—be concluded.
What would change our interpretation of a decision?
New contemporaneous evidence, omitted alternatives, missing authorization conditions and rival explanations may materially change a reconstruction. A persuasive narrative is not a substitute for evidence.
When does oversight cease to be actionable?
Formal permission is not equivalent to a feasible intervention route. Timing, execution authority, available evidence and technical safeguards have to be examined separately.
What would show that a proposed method adds no value?
If a competent existing comparator reaches the same warranted conclusions more simply and reliably, methodological redundancy is a legitimate research result—not a defect to be hidden.
Published arguments. Developing methods. Distinct forms of evidence.
Conceptual working papers and experimental protocols are not empirical validation, legal advice, regulatory certification or professional-assurance conclusions. Guest contributions are independently attributed. ID∆AC™ applied services are described separately from its academic research. The public catalogue presents selected citations; independent research records remain available through their original publisher and persistent identifiers.
ID∆AC™ Research · Independent inquiry · © 2026
Feedback and editorial correspondence ↗