Category: AI

GitHub Copilot Security

GitHub Copilot Security: The Real Risks of AI-Written Code, and How to Control Them

GitHub Copilot writes a growing share of your code, and now it also runs commands, opens pull requests and installs packages. Here are the real GitHub Copilot security risks, from prompt injection and leaked secrets to hallucinated packages, what GitHub's built-in controls cover, and seven steps to keep AI-written code under control.
9 min read
hugging face security incident

Hugging Face Security Incident: How an AI Agent Breached the Platform, and What to Fix in Your Own Pipeline

An AI agent turned one malicious dataset into access to Hugging Face's internal clusters, with no human attacker behind it. Here is what happened in the Hugging Face security incident, why it matters for every team that runs AI pipelines or AI agents, and six steps to close the same gaps in your own environment.
10 min read
AI agent security

AI agent security: know every agent, catch the risk, guard what it installs

AI agents don't just answer, they act: they call tools, hold credentials and install packages on their own. Here is how AI agent security works in four steps, from an inventory of every agent in your code to guarding what agents install.
10 min read
Xygeni at ENISE 2026 AI coding agent security

Who Decided That Release Could Ship? AI Coding Agent Security Lessons from ENISE 2026

At ENISE 2026, Xygeni CEO Jesús Cuadrado showed how one sentence in a GitHub issue led to an unauthorized release, and left CISOs with four questions about AI coding agents.
8 min read
OWASP Agentic AI Top 10

OWASP Agentic AI Top 10 Explained: ASI01–ASI10

The OWASP Agentic AI Top 10 names the ten ways AI agents fail. Here is what ASI01 to ASI10 actually mean in a real pipeline, where each risk gets in, and which ones you can stop before deploy.
11 min read
Model Extraction Attack

Model Extraction Attack: Why Inventory Comes First

A model extraction attack lets an attacker copy your model one query at a time, with no breach and no stolen file. Here is how model stealing works, and why the first defense is knowing every inference endpoint you expose.
8 min read
Agentic AI risk and blast radius

Agentic AI risk and blast radius: limit the agent, or recover faster?

Constrain the agent, recover faster, or contain it from below? Three practitioners, three answers, and one minimum they all agreed on.
13 min read
Agent harness engineering

Agent harness engineering: the model isn’t the attack surface, the harness is

Agent harness engineering turns models into useful agents. It also hands attackers a new target: the files that tell your agent what it can do.
7 min read
AI Application Security

AI Application Security: What we heard at OWASP Porto & Karlsruhe

AI application security dominated OWASP Porto and Karlsruhe. What ASPM really is, how it differs from SAST, and where AI triage and remediation help.
8 min read
AI risk management

AI risk management: two disciplines, one inventory

Four search terms, two completely different jobs. Managing the risk of AI is not the same discipline as using AI for risk management, and most organisations are staffing one while being sold the other. The interesting part is what happens when you realise they are the same programme.
9 min read
AI Risk Metrics

AI Risk: The Metrics That Tell You Whether Your Agentic AI Strategy Is Working

Every board now asks the same question: what is our AI risk? Most teams answer with adjectives. Here are the AI risk metrics that produce a number, where each one comes from, and what an agentic AI strategy has to cover before any of them mean anything.
9 min read
AI security threats

Top 10 AI security threats and how to map them

Two things get filed under the same heading and they are not the same problem. Attacks on AI systems and attacks powered by AI need different controls, different owners and different evidence. Ten threats, split into the two halves, and the one step that precedes all of them.
10 min read