APNTO FIELD NOTES / SECURITY RESEARCH

Making sense of security in the age of AI.

Practical perspectives on AI agent security, black-box testing, and software vulnerabilities—for the people building, securing, and governing modern technology.

Explore all articles
SECURITY RESEARCH
AI agentsAI testingVulnerabilities

Discover. Test. Secure.

FIELD NOTES

Latest insights

Explore practical guidance across AI security, agent governance, and software exposure.

11 published articles

AI Black-Box Testing

You Can't Read an AI Model's Mind. So Test Its Behavior.

You do not always need access to a model's weights or internal code to learn something important about its security. Black-box testing examines observable behavior under carefully designed inputs and conditions.

7 min read
Read article
Vulnerability & Exposure

Your Scanner Found 500 Vulnerabilities. Which Ones Matter?

A long vulnerability list can create the illusion of progress while leaving teams unsure where to start. Prioritization connects technical findings to exploitability, exposure, and business impact.

8 min read
Read article
AI Agent Security

The Hidden Risk Isn't the AI Agent. It's What the Agent Can Access.

An agent's model is only part of its risk profile. Connected tools, service identities, permissions, and reachable data determine what the agent may be able to do when something goes wrong.

7 min read
Read article
AI Black-Box Testing

Prompt Injection Is Only the Beginning: How to Test AI Applications

Prompt injection deserves attention, but a useful AI security assessment also tests data boundaries, tool permissions, output handling, and multi-step workflows.

9 min read
Read article
Vulnerability & Exposure

False Positives in SAST: Why More Findings Don't Mean Better Security

Static analysis can identify risky code patterns at scale, but teams still need to validate findings, understand reachability, and keep genuine issues from disappearing into alert fatigue.

8 min read
Read article
AI Agent Security

Approved, Unknown, or Unowned? A Practical Framework for AI Agent Governance

A useful governance process distinguishes what is approved, what lacks an accountable owner, and what still needs investigation—without treating missing evidence as proof of wrongdoing.

7 min read
Read article
AI Black-Box Testing

A Practical Checklist for Black-Box Testing an AI Application

Use this checklist to plan a repeatable assessment of an AI application, from defining scope and test identities to checking data boundaries, tool actions, evidence, and retesting.

9 min read
Read article
Vulnerability & Exposure

From Finding a Vulnerability to Fixing It: Closing the AppSec Loop

Detection is only the first step. A mature application security workflow validates findings, assigns ownership, guides remediation, and verifies that the fix actually worked.

7 min read
Read article
AI Agent Security

Build-Time vs. Runtime AI Agent Security: What Changes?

Build-time reviews help catch design and configuration weaknesses before deployment. Runtime controls address behavior in a live environment. Effective agent security needs both perspectives.

8 min read
Read article
Vulnerability & Exposure

CVSS Isn't the Whole Story: Prioritizing Vulnerabilities in Context

CVSS helps describe vulnerability severity, but remediation decisions also depend on exposure, exploitability, asset criticality, and the controls in place.

8 min read
Read article
Security Engineering

AI Security Is More Than Model Safety

A model can behave as intended while the surrounding application remains exposed. AI security also depends on agent permissions, connected systems, data boundaries, and the software beneath it.

8 min read
Read article
MAKE IT VISIBLE

Build a clearer picture of your security.

From AI agents and AI testing to software vulnerabilities, understand the risks that matter and decide where to act next.

Explore Our Platform