No One Is Watching the Governance Gap for Intellectual Property in AI Systems

Artificial intelligence (AI) is increasingly becoming a valuable enterprise asset, raising new questions about how organizations should protect and govern the intellectual property embedded in AI systems. Organizations need to establish governance evidence for AI intellectual property (IP) and approaches beyond the “reasonable measures” standard, write James Malackowski and Joel Lutzker for J.S. Held.

The article begins with the premise that IP law has never required absolute protection. Trade secret law, for example, focuses on whether organizations take reasonable steps to preserve secrecy. At the same time, patent and copyright practices rely on disciplined documentation for invention, ownership, authorship, and creative contribution. AI complicates these established approaches because valuable assets can span training data, model assets, algorithms, prompts, retrieval systems, infrastructure, and governance processes.

Malackowski and Lutzker introduce the concept of an AI Capital Stack, describing layered assets that require visibility before they can be protected, and propose an AI Asset Register, which provides a structured record of the organization’s AI assets. 

Governance generates an evidentiary record — including approvals, audits, and incident reports — that organizations may need during litigation, transactions, or insurance underwriting. Classifying and documenting asset lineage can establish provenance, contributors, model versions, validation activity, and material modifications. Disclosure controls are another critical component. Organizations must distinguish among consumer, enterprise, and private AI environments; establish approved-use policies; and create escalation procedures for sensitive information. 

Malackowski and Lutzker warn against AI washing, which makes public claims about capabilities that outpace actual governance maturity. They point to emerging frameworks such as the Chicago Principles for Independent AI Assurance as an evolving mechanism for verifying that governance policies operate in practice rather than existing only on paper.

Independent assurance is the next frontier for AI intellectual property and verified evidence of governance may become the most persuasive proof that an organization has exercised reasonable measures over its AI assets.