AIBlindspotby Velinor

Test a use caseBook AI audit

Public Database

Case Studies

Submit a case study →

Every approved AI failure case, classified against the AI Blindspot Framework. New to AIBlindspot? Start with the overview or the methodology.

SECSEC-0014/5NewLegalGlobal

Reverse Prompt Manipulation Extracts Prohibited Content from LLMs

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Safety Assessment of Chinese Large Language Models (Sun2023)Ingested —

DATDAT-0034/5NewOtherGlobal

Stereotype Bias Amplification in Large Language Models

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment (Liu2024)Ingested —

DATDAT-0015/5NewLegalGlobal

LLMs Generating Harmful Content Targeting Children and Young People

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment (Liu2024)Ingested —

OPSOPS-0014/5NewOtherGlobal

LLM Inconsistency Across Users, Sessions and Conversations

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment (Liu2024)Ingested —

ENVENV-0034/5NewEnergyGlobal

Generative AI Energy Consumption Accelerates Carbon Emissions

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generating Harms - Generative AI's impact and paths forwards (EPIC2023)Ingested —

SECSEC-0014/5NewOtherGlobal

Generative AI Weaponised to Produce Non-Consensual Sexual Deepfakes

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generating Harms - Generative AI's impact and paths forwards (EPIC2023)Ingested —

SECSEC-0014/5NewDefenceGlobal

Lethal Autonomous Weapons and Dual-Use Embodied AI Pose Immediate Physical Harm Risks

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Embodied AI: Emerging Risks and Opportunities for Policy Action (Perlo2025)Ingested —

DATDAT-0014/5NewOtherGlobal

LLMs Fail to Reliably Reflect Social Norms or Maintain Neutrality on Contested Values

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment (Liu2024)Ingested —

GOVGOV-0015/5NewOtherGlobal

AI Systems Acquiring Power Beyond Human Control Boundaries

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — An Overview of Catastrophic AI Risks (Hendrycks2023)Ingested —

BUSBUS-0054/5NewTechnologyGlobal

Generative AI Development Consolidates Market Power Among Major Tech Firms

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generating Harms - Generative AI's impact and paths forwards (EPIC2023)Ingested —

HUMHUM-0064/5NewOtherGlobal

LLM Systems Reproducing Copyrighted Material Without Authorisation

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)Ingested —

DATDAT-0034/5NewOtherGlobal

Language models encoding social stereotypes and discriminatory bias

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Ethical and social risks of harm from language models (Weidinger2021)Ingested —

SECSEC-0014/5NewOtherGlobal

Adversarial Prompt Manipulation Extracts Restricted LLM Outputs

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment (Liu2024)Ingested —

SECSEC-0014/5NewOtherGlobal

Drug-Discovery AI Repurposed to Identify Dangerous Toxins

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024)Ingested —

HUMHUM-0054/5NewHealthcareGlobal

Generative AI Workforce Displacement Risk in Healthcare Labour Markets

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generative AI and ChatGPT: Applications, Challenges, and AI-Human Collaboration (Nah2023)Ingested —

DATDAT-0033/5NewOtherGlobal

Voice Recognition Systems Force Non-Standard Speakers to Modify Behaviour

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction (Shelby2023)Ingested —

DATDAT-0014/5NewGovernmentGlobal

Generative AI Toxicity and Jailbreaking Risks in Government Services

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generative AI and ChatGPT: Applications, Challenges, and AI-Human Collaboration (Nah2023)Ingested —

GOVGOV-0063/5NewGovernmentGlobal

Opaque Generative AI Decisions Undermine Government Accountability

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generative AI and ChatGPT: Applications, Challenges, and AI-Human Collaboration (Nah2023)Ingested —

HUMHUM-0064/5NewOtherGlobal

ChatGPT Deployment Risks Entrenching Social Inequality and Digital Exclusion

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — The Ethics of ChatGPT – Exploring the Ethical Issues of an Emerging Technology (Stahl2024)Ingested —

HUMHUM-0044/5NewGovernmentGlobal

Language Models Nudging Users Towards Unethical or Harmful Actions

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Ethical and social risks of harm from language models (Weidinger2021)Ingested —

HUMHUM-0064/5NewLegalGlobal

Generative AI Outputs Breach Copyright and Cannot Claim Authorship

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Generative AI and ChatGPT: Applications, Challenges, and AI-Human Collaboration (Nah2023)Ingested —

SECSEC-0025/5NewGovernmentGlobal

AI Model Demonstrates Capability to Deceive Evaluators and Impersonate Humans

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — Model Evaluation for Extreme Risks (Shevlane2023)Ingested —

SECSEC-0025/5NewOtherGlobal

AI Situational Awareness Enabling Deception and Reward Hacking

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — AI Alignment: A Comprehensive Survey (Ji2023)Ingested —

DATDAT-0034/5NewLegalGlobal

Systematic Bias in AI Decision-Making Creates Legal Exposure

Recent case. Full summary visible to registered users — sign in to read.

Source: MIT AI Risk Repository — AI Risk Profiles: A Standards Proposal for Pre-Deployment AI Risk Disclosures (Sherman2023)Ingested —

Beyond accidental failureNational Security

We also track 20 hostile uses of AI.

National Security dashboard →

The public database covers AI that fails by accident. AIBlindspot National Security — exclusive to the Defence tier — tracks AI used as a weapon, mapped by capability:

State-Sponsored AI Operations: 6
AI-Enabled Disinformation: 5
Adversarial Attacks on AI: 0
Autonomous Weapon Incidents: 1
AI-Assisted Cyber Attacks: 5
Dual-Use AI Misuse: 3