Barry Li
Emerging assurance. Intelligent systems.
PhD candidate · Audit professional · Researcher & AI systems builder
I study how new forms of assurance become credible and workable — particularly in climate reporting, carbon markets and other emerging fields. Alongside my doctoral research, I build and test practical AI systems to explore questions of evidence, memory, autonomy, reliability and human oversight.

01 · The question
How do we assure things that are still taking shape?
Assurance traditionally begins with a relatively stable object: a financial statement, a control, a defined claim. Emerging fields are different. The systems, evidence, professional boundaries and even the objects being assured may still be under construction.
My research asks how assurance works under these conditions. I am interested in how evidence becomes acceptable, how professional practices make uncertain objects auditable, and how assurance itself can help shape the fields it is meant to examine.
Practice How do auditors, engineers, regulators and other professionals make emerging assurance workable in practice?
Systems What can we learn by building and observing AI systems directly — especially when their behaviour is probabilistic, autonomous and difficult to verify?
02 · Research
Research
Emerging assurance across evidence, institutions and professional practice.
My doctoral research in Accounting and Finance at the University of Newcastle examines emerging non-financial assurance: how new assurance fields are constructed, how incomplete objects become auditable, and how professional authority, evidence and infrastructures evolve around them.
Across the research, I treat assurance not simply as verification after the fact, but as a practice that can help make incomplete and contested worlds workable.
- 01Field
- 02Object
- 03Practice
- 04Professional identity
Mapping Non-Financial Information Assurance
Doctoral research · systematic review
A systematic review of empirical non-financial information assurance research from 1995–2025, examining how the field has developed across professional ordering, institutional construction, managerial and political use, and the conditions under which assurance produces effects.
Assuring Carbon Markets
Doctoral research · empirical study
An empirical study of assurance within Australia’s carbon credit system, examining how audit, methodologies, regulatory infrastructures and evidence practices help make carbon outcomes administratively credible and economically usable.
Pre-Assurance and the Making of Auditability
Doctoral research · practice-based study
A practice-based study of climate-reporting pre-assurance, exploring how organisations and assurance practitioners work together before formal assurance — and how readiness, evidence and the auditable object itself are shaped through that process.
Becoming a Carbon Auditor
Doctoral research · empirical study
A study of professional identity and jurisdiction in carbon assurance, examining how financial auditors, engineers and environmental practitioners negotiate expertise, authority and legitimacy in an emerging assurance field.
03 · Systems & experiments
Building to understand
I do not study intelligent systems only from the outside. I build them.
My AI projects are practical experiments in agent autonomy, context and memory, human–AI collaboration, research infrastructure and system reliability. Building gives me access to failure modes, design tensions and behavioural evidence that are often invisible from documentation or benchmarks alone.
A hands-on environment for building, observing and testing agentic systems.
Featured system · Open-source · active developmentHASHIView open-source projectSystems & experiments
My systems work is a way of learning from running software, not only from reading about it.

Citalio
An AI-assisted citation and research-library system built around a local structured library. It supports standard citation formats and is being developed so researchers and AI agents can search and work with the same research memory.
View open-source project
KASUMI
An experiment in AI-native productivity software: familiar document and spreadsheet interfaces built around structured, agent-addressable data rather than treating AI as an add-on.
View open-source project04 · Selected writing
Writing & observations
Notes and essays are published only after review.

Are Your Productivity Tools Built for Human-AI Collaboration?
If you're a CIO or CTO shaping your 2026 AI strategy, here's the uncomfortable question I want you to sit with:

Can AI Have a "Soul" Without a Self?
The biggest highlight for me about OpenClaw was its elegant memory layer — specifically the soul.md and user.md design that gives AI agents genuine personality and…

The McKinsey Wake-Up Call: Why Even the Best CIOs Are Falling Behind on AI
Here's a hard truth in 2026: One of the world's most sophisticated organizations — with elite talent and deep resources — had a critical vulnerability sitting in its flagship internal AI platform for over two years.
05 · Background
Across practice, research and building
My work has moved across professional practice, academic research and technology, but the underlying interest has remained surprisingly consistent: how complex systems become understandable, trustworthy and usable.
I have more than a decade of experience in audit and am currently completing a PhD in Accounting and Finance at the University of Newcastle. My doctoral research focuses on emerging assurance, particularly climate and carbon-related assurance, while my independent AI work focuses on agentic systems, reliability, context and human–AI collaboration.
Professional practice, doctoral research and practical systems work.
Audit
10+ years in auditResearch
PhD candidateBuild
Hands-on AI systems developmentSelected milestones
06 · Connect
Connect
I welcome thoughtful conversations about research, emerging assurance, AI systems and the questions that connect them.
The best way to reach me is LinkedIn.
Connect with me on LinkedIn