Model robustness
When do models become unreliable?
Investigating how changing conditions expose failures that aggregate scores can miss.
Related workOur research connects model behaviour, data quality, and evaluation methods. We start with the conditions under which a system can fail, then ask what evidence would reveal it.
Questions we investigate
Data quality
What does the data leave out?
Examining missing examples, inconsistent records, and the relationship between data and model behaviour.
Related workEvaluation & assurance
What evidence supports a deployment decision?
Connecting test criteria, observed outcomes, and risk records so that decisions can be examined.
Related workLab notes
Original observations and attributed reading from the wider field of AI evaluation.
Reading note
Stanford HAI AI Index 2026: What Executives Should Track
Stanford HAI's 2026 AI Index is a core research reference for AI capability, adoption, investment, policy, and governance trends.
Source: hai.stanford.edu ↗
AI Governance
Reading note
MIT AI Risk Repository: A Practical Taxonomy for Risk Registers
MIT's AI Risk Repository helps governance teams sort AI risks into categories that work for registers, controls, and executive review.
Source: airisk.mit.edu ↗
AI Risk Management
Reading note
CSET AI Governance Research: Enterprise Signals to Watch
CSET policy research helps enterprises understand AI governance, security, capability trends, and institutional readiness.
Source: cset.georgetown.edu ↗
AI Governance
Reading note
RAND AI Research: Risk and Policy Signals for Enterprise Leaders
RAND AI research helps executives assess AI through risk, security, policy, and operational consequence lenses.
Source: rand.org ↗
AI Risk Management