algorithmic-auditing

Code for the paper "ClipMind: A Framework for Auditing Short-Format Video Recommendations Using Multimodal AI Models"

Depth-tracked regulatory audit primitives for privacy-preserving AI audits with signed envelopes and TenSEAL CKKS support.

Persona-driven agents that build real signed-in YouTube watch histories, for auditing algorithmic personalization

The comprehensive list of AI governance tools: 130+ platforms and resources for AI risk, compliance, model governance, audit, fairness, privacy, and responsible AI.

Open-source statistical engine for auditing AI employment decision systems: adverse impact analysis under the Uniform Guidelines (29 CFR 1607), name-swap bias testing, and model drift detection

Auditing a 0.928-AUC loan-approval model for fairness. Accuracy was never the problem: employment status carries the disparity, and a default 0.5 threshold quietly penalises students and unemployed borrowers. NYU Responsible Data Science.

Pilot audit of cultural bias in vision–language models: LLaVA-NeXT associates South Asian subjects with domestic roles 9× more often, and exoticises South Asian cultural events 14× more.

Systematic safety audit of a ResNet-50 diabetic retinopathy classifier using the Medical Algorithmic Audit framework (Liu et al., 2022). Included error analysis, subgroup testing, adversarial robustness, and FMEA risk scoring on the APTOS 2019 dataset.

Does an LLM told to favour a demographic group say so in its written justification? Collection and analysis harness for a four-arm crossed study of disclosure in LLM résumé screening.

Bias audit toolkit reproducing the recruitment AI case from Activity 6 — Unidad 3, Electiva II Inteligencia Artificial Avanzada. Empirical analysis with reweighing mitigation, Model Card documentation, and APA 7 references.