CAST is an open framework for responsible AI design and engineering. It offers design heuristics and patterns, and RAI recommendations through generative features and online content.

ADMIT is a research tool within a broader methodological framework combining quantitative and qualitative strategies to identify, analyse, and mitigate social implications associated with automated decision-making systems while enhancing their potential benefits. It supports comprehensive assessments of sociotechnical impacts to inform responsible design, deployment, and governance of automation technologies.

HiddenLayer’s AISec Platform is a GenAI Protection Suite purpose-built to ensure the integrity of AI models throughout the MLOps pipeline. The platform provides detection and response for GenAI and traditional AI models to detect prompt injections, adversarial AI attacks, and digital supply chain vulnerabilities.

This tool provides a comprehensive risk management framework for frontier AI development, integrating established risk management principles with AI-specific practices. It combines four key components: risk identification through systematic methods, quantitative risk analysis, targeted risk treatment measures, and clear governance structures.

An open-source Python library designed for developers to calculate fairness metrics and assess bias in machine learning models. This library provides a comprehensive set of tools to ensure transparency, accountability, and ethical AI development.

Behavior Elicitation Tool (BET) is a complex-AI system that systematically probes and elicits specific behaviors from cutting-edge LLMs. Whether for red-teaming or targeted behavioral analysis, this automated solution is Dynamic Optimized and Adversarial (DAO) and can be configured to test the robustness precisely and help to have a better control of the AI system.

MLPerf Client is a benchmark for Windows and macOS, focusing on client form factors in ML inference scenarios like AI chatbots, image classification, etc. The benchmark evaluates performance across different hardware and software configurations, providing command line interface.

Continuous proactive AI red teaming platform for AI and GenAI models, applications and agents.

Vectice is a regulatory MLOps platform for AI/ML developers and validators that streamlines documentation, governance, and collaborative reviewing of AI/ML models. Designed to enhance audit readiness and ensure regulatory compliance, Vectice automates model documentation, from development to validation. With features like automated lineage tracking and documentation co-pilot, Vectice empowers AI/ML developers and validators to work in their favorite environment while focusing on impactful work, accelerating productivity, and reducing risk.

Dioptra is an open source software test platform for assessing the trustworthy characteristics of artificial intelligence (AI). It helps developers on determining which types of attacks may impact negatively their model's performance.

Responsible AI (RAI) Repairing Assistant

The Algorithm Charter for Aotearoa New Zealand is a set of voluntary commitments developed by Stats NZ in 2020 to increase public confidence and visibility around the use of algorithms within Aotearoa New Zealand’s public sector. In 2023, Stats NZ commissioned Simply Privacy to develop the Algorithm Impact Assessment Toolkit (AIA Toolkit) to help government agencies meet the Charter commitments. The AIA Toolkit is designed to facilitate informed decision-making about the benefits and risks of government use of algorithms.

ISACA’s Digital Trust Ecosystem Framework (DTEF) offers enterprises a holistic framework that applies systems thinking – the notion that a change in one area can have an impact on another area – across an entire organisation.

NayaOne is a Sandbox-as-a-Service provider to tier 1 financial services institutions, world-leading regulators, and governments. This sandbox is designed to address key concerns in AI deployment by providing a single environment where AI can be evaluated and procured while also enabling collaboration and access to world-leading tools.

LangBiTe is a framework for testing biases in large language models. It includes a library of prompts to test sexism / misogyny, racism, xenophobia, ageism, political bias, lgtbiq+phobia and religious discrimination. Any contributor may add new ethical concerns to assess.

PRIDAR (Prioritization, Research, Innovation, Development, Analysis, and Review) A Risk Management Framework

A Python package for identifying 42 kinds of animals, training custom models, and estimating distance from camera trap videos

Code for our nips19 paper: You Only Propagate Once: Accelerating Adversarial Training Via Maximal Principle

Wild Me's first product, Wildbook supports researchers by allowing collaboration across the globe and automation of photo ID matching

Explainability for Vision Transformers