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.
Reading note · AI Governance
Source: hai.stanford.edu
HephaKnot is an applied AI research lab. We test models in the conditions they run in, check the data behind them, and keep the evidence for each result.
- Model robustness
- When do models become unreliable?
- Data quality
- What does the data leave out?
- Evaluation & assurance
- What evidence supports a deployment decision?
Our focus
A model can pass a benchmark. The world keeps changing.
Different weather. Missing examples. An unfamiliar document. We investigate how changes in data and conditions affect model behaviour, and what evidence helps us understand its limits.
Why we do this workOur approach
Start with a question. Follow the evidence.
Research areas
Experiments
Define · Compare · Examine
Lab notes
From the notebook.
Observations, analysis and reading that inform our work on AI evaluation.
All notes & readingEU AI Act Regulatory Framework: 2026 Official Reference
The European Commission's AI Act policy hub remains a primary official reference for risk-based AI obligations and enterprise governance planning in 2026.
Reading note · AI Governance
Source: digital-strategy.ec.europa.eu
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.
Reading note · AI Risk Management
Source: airisk.mit.edu
Knotest
Tools for the work
Knotest brings model tests, data checks, and their supporting evidence into one workspace.
