Identify the architecture
Map the actors, mechanisms, boundaries, dependencies, and flows that constitute the system.
Founder-led independent research institution
Understanding and governing the causal architectures that generate technological, institutional, and societal outcomes.
Causal Architectonics examines the structures, constraints, incentives, and feedback systems from which complex outcomes emerge.
Explore the researchResearch institution·Foundational work in progress
Bjorn Bjornsvik·Principal Investigator
01
Approach
Outcomes are visible. Their generating structures often are not.
Complex failures are frequently addressed at the level at which they become observable: an incident, a harmful decision, a market distortion, or an institutional breakdown. Causal Architectonics begins further upstream.
The research asks which arrangements of authority, optimization, information, incentives, and constraint made an outcome possible, distributed responsibility, or rendered failure increasingly likely.
Trace causal production
Examine how local actions, feedback loops, and optimization pressures combine to produce outcomes.
Intervene at the source
Develop controls and institutional arrangements directed at generating conditions rather than symptoms alone.
02
Current research
Research portfolio
Two programmes and one standalone jurisprudential work are presently open to public view. They retain distinct research and publication positions.
Independent research programme
Machine Bylaws
Agentic AI Control Framework
An independent research programme examining how increasingly autonomous artificial-intelligence systems create new questions of authority, accountability, oversight, and control.
Two foundational publications released · Wider programme under developmentView research
Research and educational programme
Reading the Universe
A Developing Book and Research Programme
A developing book and research programme examining how mathematical and scientific frameworks are constructed, interpreted, evidenced, and inherited. The work approaches mathematics as a language of structured description and seeks to make the architecture beneath scientific explanation more visible.
Manuscript in development · Future publisher engagement plannedView research
Standalone jurisprudential research work
Artificial Agency Without Personhood
Toward a Jurisprudence of Autonomous AI
A complete standalone jurisprudential work examining how law can attribute conduct and allocate responsibility when autonomous systems select consequential means without becoming legal persons.
Version 1.0 published · Preprint, not peer reviewedView research
Additional programmes will be introduced only when they reach an appropriate stage for publication, funding, or public engagement. View the wider research landscape
03
Publications
Publication register
Three DOI publications.
Publications are registered here only after formal public release and linked to the research context to which they belong.
Delegated Authority Without Accountability
The Governance Discontinuity Created by Agentic AI
Bjorn Bjornsvik · 2026 · Zenodo
A foundational examination of the governance discontinuity created as AI systems acquire increasingly autonomous and consequential capabilities.
Read the DOI publicationMachine Bylaws: Constitutional Governance for Non-Human Operational Actors
Constitutional Conditions for Governable Non-Human Agency
Bjorn Bjornsvik · 6 September 2026 · Zenodo
A constitutional account of the conditions required for non-human operational agency to remain governable under delegated authority.
Read the DOI publicationArtificial Agency Without Personhood
Toward a Jurisprudence of Autonomous AI
Bjorn Bjornsvik · 10 September 2026 · Zenodo
A comparative, function-oriented examination of legal attribution and responsibility for consequential conduct produced by autonomous objective-directed systems.
04
About
Institutional position
The institutional home of an independent research portfolio.
Causal Architectonics is a founder-led independent research institution and the institutional home of its research portfolio. It is presently unincorporated and led by a single Principal Investigator.
Its mandate is aspirational but substantive: to develop rigorous, transferable methods for understanding and governing the architectures that produce consequential outcomes across complex systems.