[Paper Review] Outlining Traceability: A Principle for Operationalizing Accountability in Computing Systems

Joshua A. Kroll|arXiv (Cornell University)|Jan 23, 2021

Ethics and Social Impacts of AISocial Sciences141 references56 citations

TL;DR

This paper proposes traceability as a foundational principle for operationalizing accountability in computing systems by linking system design, development, and operation to normative goals like fairness and safety. It systematizes traceability requirements, maps existing tools and techniques, and identifies critical gaps between policy mandates and practical implementation.

ABSTRACT

Accountability is widely understood as a goal for well governed computer systems, and is a sought-after value in many governance contexts. But how can it be achieved? Recent work on standards for governable artificial intelligence systems offers a related principle: traceability. Traceability requires establishing not only how a system worked but how it was created and for what purpose, in a way that explains why a system has particular dynamics or behaviors. It connects records of how the system was constructed and what the system did mechanically to the broader goals of governance, in a way that highlights human understanding of that mechanical operation and the decision processes underlying it. We examine the various ways in which the principle of traceability has been articulated in AI principles and other policy documents from around the world, distill from these a set of requirements on software systems driven by the principle, and systematize the technologies available to meet those requirements. From our map of requirements to supporting tools, techniques, and procedures, we identify gaps and needs separating what traceability requires from the toolbox available for practitioners. This map reframes existing discussions around accountability and transparency, using the principle of traceability to show how, when, and why transparency can be deployed to serve accountability goals and thereby improve the normative fidelity of systems and their development processes.

Motivation & Objective

• To operationalize accountability in computing systems through the principle of traceability.
• To identify and systematize the technical, procedural, and governance requirements for achieving robust traceability.
• To bridge the gap between policy-level traceability mandates and the practical tools available to implement them.
• To clarify how traceability supports accountability beyond mere transparency by linking system behavior to design decisions and values.
• To provide a structured framework for practitioners and policymakers to assess and improve system traceability.

Proposed method

• Analyzes policy documents and AI principles from around the world to extract recurring traceability requirements.
• Maps traceability requirements to existing technologies, tools, and development practices such as version control, provenance tracking, and audit logging.
• Distinguishes traceability from explainability by emphasizing causal, contrastive, and selective explanations of design and operational decisions.
• Proposes a framework that connects system outputs to development processes, data sources, and governance goals.
• Identifies four key categories of gaps: technology limitations, tooling deficiencies, process integration challenges, and stakeholder understanding.
• Uses a systematic review of standards and policy guidance to derive a taxonomy of traceability needs.

Experimental results

Research questions

• RQ1 How can traceability be systematically operationalized in software and AI systems to support accountability?
• RQ2 What specific technical and procedural requirements are necessary for achieving meaningful traceability in practice?
• RQ3 How does traceability differ from transparency and explainability in supporting normative governance goals?
• RQ4 What gaps exist between policy-level traceability mandates and the current state of available tools and practices?
• RQ5 In what ways can traceability improve the normative fidelity of computing systems to values like fairness, safety, and non-discrimination?

Key findings

• Traceability is a more concrete and measurable principle than abstract values like fairness or transparency, making it a practical lever for accountability.
• Existing tools such as version control, provenance tracking, and audit logging can support traceability but are often underutilized or misapplied.
• A critical gap exists between policy mandates requiring traceability and the actual implementation capabilities of development teams.
• Traceability requires more than mechanical logging—it demands human-understandable records of design decisions, purpose, and value trade-offs.
• The principle of traceability enables accountability by connecting system behavior to its origins, thus supporting responsibility and normative assessment.
• The paper identifies four major categories of implementation gaps: technology, tooling, process integration, and stakeholder comprehension.