# Eticas.ai

> Eticas.ai is an independent AI evaluation and assurance firm that has audited AI systems in production since 2012, testing them against bias, privacy, reliability, governance, security, environmental impact, transparency, and agentic-risk criteria.

Eticas.ai runs structured, evidence-based evaluations of deployed AI systems — not policy checklists or self-reported surveys — and delivers scored findings, monitoring, and recommendations. The company also operates Eticas Foundation, a nonprofit dedicated to algorithmic accountability, and maintains the Eticas AI Risk Taxonomy, a versioned, publicly addressable vocabulary of AI risk categories.

## Core pages
- [What We Do](https://www.eticas.ai/what-we-do): Eticas's AI System Risk Management Process — a four-step cycle of Forward Deployment advisory, a First Evaluation that produces a board-readable risk picture and a shareable "Eticas.ai Seal" scorecard, Automated Monitoring with drift alerts, and recurring New Evaluations as systems change. Evaluations cover the full system — data flows, model, business rules, UX, human oversight, and real-world outcomes — typically take days to a couple of weeks, and are designed to complement (not replace) GRC processes and standards like ISO/IEC 42001 and the EU AI Act.
- [Our Work](https://www.eticas.ai/our-work): Case studies organized by the three system types Eticas evaluates — expert & predictive systems, LLM-based systems, and agentic systems — and by industry (healthcare, education, HR & hiring, public sector), each with the specific risks tested and real findings from completed evaluations.
- [The Platform](https://www.eticas.ai/the-platform): How Eticas evaluates AI systems with human experts empowered by technology — a repeatable four-step process (Scope, Evidence, Evaluate, Deliver then monitor) and an open methodology that chains Risk Concept (what can go wrong), Measurement (how it is detected), and Grade (how severe it is). Findings are graded A to E rather than pass/fail, with pattern flags (systemic, focal, isolated), and the taxonomy is mapped to the EU AI Act, ISO/IEC 42001, and NIST AI RMF. Deliverables include a Full Report, Evaluation Brief, Score, and Badge, with optional ongoing monitoring. See also the [Eticas AI Risk Taxonomy](https://taxonomy.eticas.ai/risk/) and the [methodology paper](https://arxiv.org/abs/2607.02201).
- [About Us](https://www.eticas.ai/about-us): Founded in 2012 as the first company dedicated to AI auditing; 200+ evaluations in high-risk contexts across four continents. Covers founder Dr. Gemma Galdon-Clavell's background (advisor to the European Commission, OECD, UN, and World Economic Forum; President of the International Association of Algorithmic Auditors) and the company's public, GitHub-versioned Responsible AI Use Policy governing its own internal AI use.
- [Contact Us](https://www.eticas.ai/contact-us): Get in touch for an evaluation, a speaking engagement, press enquiries, or to join the team.

## AI Risk Taxonomy
- [Eticas AI Risk Taxonomy (v2.0.0)](https://taxonomy.eticas.ai): Versioned, URI-addressable controlled vocabulary of 8 AI risk categories — Bias & Fairness, Privacy & Confidentiality, Reliability, Governance, Security & Misuse, Environmental Impact, Transparency & Explainability, Agentic AI.

## Knowledge Hub — Papers & Guides
- [The Eticas Risk Taxonomy Is Now Part of IBM's Risk Atlas Nexus](https://www.eticas.ai/knowledge/risk-atlas-nexus-eticas-taxonomy): Announcement about IBM Research adding the Eticas Risk Taxonomy to IBM’s Risk Atlas Nexus, its open-source toolkit for AI governance. 
- [Why Aggregate Metrics Miss Bias in AI Hiring](https://www.eticas.ai/knowledge/barcelona-activa-ai-hiring-audit-methodology): Methodology behind a stage-by-stage evaluation of a hiring AI system that found five disparities hiding underneath.
- [The Eticas AI Risk Taxonomy: Operationalizing AI Evaluations with Measurable Proof](https://www.eticas.ai/knowledge/ai-risk-taxonomy-operationalizing-audits): How the taxonomy translates AI principles into measurable audit criteria.
- [Responsible AI in Healthcare: A Practical Guide to Risk, Oversight, and Independent Assurance](https://www.eticas.ai/knowledge/responsible-ai-in-healthcare-where-risk-concentrates): Where risk concentrates in healthcare AI and how to oversee it.
- [Equitable AI for Outcomes](https://www.eticas.ai/knowledge/equitableai-for-outcomes-2026): Framework for evaluating AI systems for equitable outcomes.
- [Responsible AI for Wellbeing Apps](https://www.eticas.ai/knowledge/putting-responsible-ai-at-the-heart-of-two-groundbreaking-digital-health-applications): Ethics and algorithmic review of two digital wellbeing apps, identifying opportunities to strengthen data transparency, accessibility, and inclusivity.
- [When AI Enters the Classroom: Why Real-World Evidence Matters](https://www.eticas.ai/knowledge/when-ai-enters-the-classroom-why-real-world-evidence-matters)

## Knowledge Hub — Case Studies
- [Evaluation of an Algorithmic Hiring System: A Public-Sector Case Study](https://www.eticas.ai/knowledge/auditting-a-public-service-recruiting-system): A five-year, post-deployment fairness and bias evaluation of a high-risk, EU AI Act–classified hiring platform used by a European public employment agency. Found systematic shortlisting disparities by gender, age, education level, and national origin, including near-total exclusion of candidates aged 55 and over.
- [Auditing an AI-Enabled Career Advisory Platform](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform): Post-deployment audit of an AI-enabled career advisory platform for students aged 13+, confirming a safe, reliable, and age-appropriate deployment with targeted recommendations for improvement.
- [Deep Learning for Social Services](https://www.eticas.ai/knowledge/deep-learning-for-social-services)
- [Detecting and Preventing Bias in AI Hiring Systems](https://www.eticas.ai/knowledge/detecting-and-preventing-bias-in-ai-hiring-systems-research-project-presents-new-tools): Part of the FINDHR project, introducing practical tools, guidelines, and auditing frameworks to reduce discrimination in AI-assisted hiring across Europe.
- [Auditing an AI-Based Cybersecurity Application](https://www.eticas.ai/knowledge/auditing-an-ai-based-cybersecurity-application)
- [Fairness in Public-Sector AI: Allegheny County Homelessness Risk Decisions](https://www.eticas.ai/knowledge/allegheny-county-audit-for-equitable-homelessness-risk-decisions)
- [Audit of an AI-Based Wellbeing Support Application](https://www.eticas.ai/knowledge/audit-for-an-emotional-ai-companion): Evaluation of an AI wellbeing companion's safety, fairness, and privacy across sensitive real-world scenarios, sharpening crisis-response protocols and reducing subtle bias.

## Knowledge Hub — Press, Interviews & Talks
- [WIRED Trusts Eticas' Expertise in UK Investigation](https://www.eticas.ai/knowledge/iwired-june2026): Bias-testing methodology cited in a WIRED investigation into UK policing AI.
- [IAPP AI Governance Global Europe](https://www.eticas.ai/knowledge/iapp-june2026)
- [CEO Interview by InnovEU – The EU Project Chronicles](https://www.eticas.ai/knowledge/innoveu-april2026)
- [Bold Journey Magazine Interviews Gemma](https://www.eticas.ai/knowledge/boldjourney-interview-march2026)
- [CEO Interview by Armilla.ai](https://www.eticas.ai/knowledge/armila-interview-march2026)
- [Frecuencia CEO Podcast (Spanish)](https://www.eticas.ai/knowledge/frecuenciaceopodcast-june2026)
- [CEO Interview by Deusto University (Spanish)](https://www.eticas.ai/knowledge/deustouniversity-october2023)
- [Gemma Galdon-Clavell Keynote at MAICON 2022](https://www.eticas.ai/knowledge/maicon-keynote-august-2022)
- [Eticas Co-Founds the International Association of Algorithmic Auditors (IAAA)](https://www.eticas.ai/knowledge/cofoundingiaaa-dec2023)

## Knowledge Hub — Glossary
- [Algorithmic Accountability](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh-wg5ap-yl838-3dgnp-lrmf6-t5nkk)
- [Fairness Testing](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh-wg5ap-yl838-3dgnp-lrmf6)
- [Model Drift](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh-wg5ap-yl838-3dgnp)
- [AI Explainability](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh-wg5ap-yl838)
- [Post-Deployment Monitoring](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh-wg5ap)
- [AI Risk Assessment](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h-944wh)
- [EU AI Act](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83-zjh6h)
- [Socio-Technical Auditing](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp-e6j83)
- [AI Governance](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat-d5lmp)
- [Algorithmic Bias](https://www.eticas.ai/knowledge/auditing-ai-enabled-career-advisory-platform-9bdat)


## Optional
- [Press Kit](https://www.eticas.ai/press-kit)
- [LinkedIn](https://www.linkedin.com/company/eticasai/)
- [GitHub](https://github.com/Eticas-AI)