Inspectable AI for Science: A Research Object Approach to Generative AI Governance
Organizations: CISPA, Helmholtz Center for Information Security · DATA61,CSIRO · Data61, CSIRO · Macquarie University
Abstract
This paper introduces AI as a Research Object (AI-RO), a paradigm for governing the use of generative AI in scientific research. Instead of debating whether AI is an author or merely a tool, we propose treating AI interactions as structured, inspectable components of the research process. Under this view, the legitimacy of an AI-assisted scientific paper depends on how model use is integrated into the workflow, documented, and made accountable. Drawing on Research Object theory and FAIR principles, we propose a framework for recording model configuration, prompts, and outputs through interaction logs and metadata packaging. These properties are particularly consequential in security and privacy (S&P) research, where provenance artifacts must satisfy confidentiality constraints, integrity guarantees, and auditability requirements that generic disclosure practices do not address. We implement a lightweight writing pipeline in which a language model synthesizes human-authored structured literature review notes under explicit constraints and produces a verifiable provenance record. We present this work as a position supported by an initial demonstrative workflow, arguing that governance of generative AI in science can be implemented as structured documentation, controlled disclosure, and integrity-preserving provenance capture. Based on this example, we outline and motivate a set of necessary future developments required to make such practices practical and widely adoptable.
Figures & tables
| Governance Dimension | Traditional Research Objects | AI Research Objects |
|---|---|---|
| Ontological status | Digital scientific artefact | Digital scientific artefact |
| Typical examples | Datasets, software, workflows | AI use logs, AI agents |
| Versioning | Dataset/software versions | Model versions, checkpoints |
| Configuration | Parameters, scripts | Temperature, Top-p, Max tokens |
| Provenance | Data sources, pipelines | Prompt logs |
| Reproducibility | Full: Re-run workflow | Limited: Re-run model with same context |
| Artifact | Captured Information | Governance Function | Primary User |
|---|---|---|---|
| AI-ROIC | Model identity, configuration parameters, intended use, released artifacts, limitations, and oversight description | Provides a high-level inspection interface summarizing how AI was used and governed; supports transparency and interpretability | Authors, Readers, reviewers |
| Prompt template | Constraints, task framing, allowed inputs, and generation rules | Documents how model behavior was bounded; prevents hidden or unconstrained synthesis | Authors, reviewers |
| Interaction/invocation logs | Timestamps, parameters, cryptographic hashes, prompt–response linkage | Preserves provenance and enables auditability without exposing sensitive raw content | Authors, Auditors, editors |
| Human-authored input bundle | Verified references, structured notes, contribution framing | Establishes human grounding of claims and prevents fabricated or unsupported synthesis | Authors, reviewers |
| Generated draft artifact | Model-produced synthesis text and intermediate reasoning structure | Exposes how AI influenced writing; enables comparison with final authorial decisions | Auditors, researchers |
| Human edit trace | Tracked revisions distinguishing AI suggestions from author decisions | Demonstrates human verification and responsibility over final claims | Reviewers, auditors |
AI Governance
Artificial intelligence governance studies how laws, institutions, and technical controls shape AI development, deployment, accountability, and social effects.
See all 379 papers