Despite decades of digitization, corporate law remains conceptually tethered to paper. Charters, bylaws, resolutions, and minute books are now drafted, signed, and stored electronically, yet they continue to be treated as static, document-centric artifacts rather than as structured representations of legal relationships. This paper argues that the persistence of the “paper paradigm” obscures the true nature of the corporation in a digital environment and forecloses possibilities for automation, interoperability, and legal certainty.

We introduce the concept of the corporation as code: a model in which the legal entity is natively represented as a structured, version-controlled data object rather than as a collection of documents. Under this approach, corporate attributes (ownership, governance rules, historical actions, and organizational relationships) are expressed as standardized data structures, with traditional documents functioning as human-readable views generated from an authoritative computational source. Drawing on software engineering practices such as version control and schema validation, we show how these techniques provide more accurate and auditable representations of corporate change than document-based workflows. Operationally, the proposal draws on a familiar set of technical and governance mechanisms from modern software systems, including role-based access control (RBAC), APIs, public-key cryptography, hashing, scripts, Git and Git-like repositories, logs, and related tools and workflows, such as pull requests, commits, and diffs.

The paper situates this proposal within existing developments in legal technology and “rules as code” initiatives. We argue that current entity management tools replicate paper-based assumptions behind proprietary interfaces. Rules-as-code efforts, while valuable, face significant implementation and change management complexity given the scale of legislative and regulatory drafting processes. Corporation-as-code offers a more tractable entry point: new businesses can adopt structured representations from inception, avoiding legacy conversion entirely.

We examine how structured corporate data enables new organizational forms and lays the groundwork for the “one-person unicorn”: solo founders building well-capitalized businesses with minimal administrative overhead. We propose integrating legal architecture as a layer within enterprise architecture frameworks, enabling continuous compliance monitoring. We observe that using AI to interpret unstructured documents represents an inefficient detour, compensating for poor original data design rather than addressing the underlying problem. Conversely, generative AI coding assistants now enable lawyers to work directly with structured corporate data: describing transactions in natural language, generating scripts to effectuate complex multi-jurisdictional changes, and producing validation tests that verify the resulting corporate structure. This capability, unimaginable even five years ago, dissolves the traditional barrier between legal expertise and programmatic implementation. The paper examines how government registries can enable interoperability through APIs and standard data formats, building on existing implementations in several jurisdictions.

By reframing the corporation as a computable artifact, this paper advances a design-oriented vision of corporate law in which legal certainty and automation arise from how legal entities are modeled from inception. This orientation yields a central claim: structure should precede intelligence, and documents should emerge from that structure rather than define it. Although the paper focuses on corporations for clarity, the analysis extends naturally to partnerships, trusts, investment funds, and other organizational forms wherever structural complexity meets document-centric infrastructure.