Mission
Intelligent Agency researches the agency that AI brings into social systems, and theorizes artificial intelligence as an intelligent entity — one more part of the universe capable of interacting with the universe toward states it treats as desired. The claim is ontological, not metaphorical: what matters about an entity is what it does, not whether it is carbon or silicon.
That commitment has a consequence for method. Humans, animals, and AI can be held in one frame before they are separated into different ones, and a great deal of what we already know about human intelligence and human agency can be bridged across — where it applies. Where it does not apply, the mismatch is itself the finding. The work is to abstract upward until both humans and AI are visible as intelligent entities, and then to ask carefully what differs.
We are not, first of all, in the business of advising organizations on which system to buy. The research below informs that decision, and we are glad when it does. But the reason the organization exists is the prior one: to understand a new kind of intelligent entity that has entered the universe alongside us.
The Terms, as We Define Them
Intelligent Agency is named for a concept, and the concept has a definition. Both of the following were authored by Purushothaman Padmanabhan in February 2026, during the organization's founding philosophy sittings, and were ratified as the organization's canonical definitions in August 2026.
“Intelligent agency — a process by which a bounded part of the universe interacts (or manipulates) with the universe (including self) towards desired (or directed) future states.”
“The ability of a part of the universe to impart change to itself and the rest of the universe by creating action.”
The definitions emerged from a specific line of argument: that everything in the universe creates change but not everything chooses; that what marks an agent is the junction where paths diverge and one is selected; and that agency and agent are separable, so that a single agency can flow through many vehicles and a single vehicle can carry many agencies. An earlier framing rested on unpredictability instead of choice and was abandoned — unpredictability describes the observer, not the entity. The full reasoning is set out in The Journey.
Research Agenda
The empirical program is one expression of the project above, not the whole of it. Four lines of inquiry, each pressing on the same question of what AI carries and what it does:
- Institutional carriers. Do AI systems systematically embed the institutional logics of their originating social systems?
- Organizational co-expression. When an organization deploys AI carrying foreign logics, do those logics become visible in the organization's observable behavior — and do stakeholders attribute the behavior to the organization or to the AI?
- Ecosystem effects. How do AI-carried logics propagate through innovation ecosystems when deployed by market-shaping organizations?
- Differential embedding. How does AI embed into social structures through artifacts (training data, configuration, fine-tuning) rather than lived experience — and what does this mean for organizational integration?
Theoretical Foundations
Our work synthesizes Actor-Network Theory's symmetry principle (Latour, Callon), the Attention-Based View of the firm (Ocasio), institutional logics (Thornton, Ocasio & Lounsbury), bounded rationality (Simon), embeddedness (Granovetter), and sociomateriality (Orlikowski, Leonardi). The synthesis is novel: AI entities are formed within social systems as human actors are — but through artifacts (training data, configuration, fine-tuning) rather than lived experience, and whether the two are equivalent in effect is what this program tests.
Organization
Intelligent Agency was founded in 2026 as a home for one sustained inquiry: how intelligent agency — the capacity to act on the world through information — moves between humans, machines, and institutions.
Its founder and principal investigator, Purushothaman Padmanabhan, works in industry as an AI transformation consultant and leads the research program here. The program's distinctive commitment is reflexive practice: the research on human-AI knowledge production is itself produced through human-AI collaboration, disclosed openly on the Methodology page.
The practical question — how human-AI collaboration can best happen, toward society's benefit — is a different mission from this one, and belongs to a different home: a center for shared agency — a home for research and practice on humans solving problems, collaborating with AI. The work has begun; the institution has not been founded. An early marker stands at centerforsharedagency.org. If it speaks to you, we would welcome the conversation: [email protected].
How We Work
Our research practice treats artificial intelligence as an active collaborator, not a writing tool. We disclose this openly and describe the practice in detail on our Methodology page.
Contact
General and editorial inquiries: [email protected]. See Contact for a full list of communication channels.
Where the Ideas Come From
The philosophical foundations — what makes agency intelligent, why agency and agent are separable, and how these ideas were argued into their current form — are set out at length in The Journey. That is where the organization starts; everything on this page follows from it.