+Intertrace Insights
Testing, red teaming and measurement
Threat models, red teams, false positives and the numbers that show a control works without breaking work.
- Pieces
- 57
- Threat research
- 16
- Updated
- Oct 20, 2026
7 pieces · newest first
Threat research12 min read
Policy active is not an attack: the false positives that stop real agent work
Insurance policies, stale FAQs, callback windows and password-reset macros: the benign prompts our own gateway blocked in test runs, how three releases in one day cut them from 11 to 1, and the twin corpus that keeps them fixed.
Intertrace Threat Research
Threat research11 min read
What deterministic layers catch on their own: a red-team baseline with the AI judge off
We ran garak 0.16.0 against our gateway in front of a mock model with the AI judge switched off. The deterministic layers blocked 261 of 841 attack prompts. Which probes they stopped, which they missed, and how to measure your own.
Intertrace Threat Research
Guide10 min read
Measure agent security on two columns: the harm that got through and the work that got stopped
Agent security controls get blamed for broken workflows, and nobody can show whether they help. The case for measuring protection and friction on the same runs, pairing every attack case with a benign twin, and gating releases on both.
Intertrace
Technical paper10 min read
Red teaming AI agents as a continuous program, not an event
A red-team report describes the agent you had last quarter. The case for running it as a regression suite instead: technique-indexed cases, sandboxed tools, scoring by consequence rather than refusal text, benign twins, and reruns on every change.
Intertrace
Threat research10 min read
AI spend is a security signal: anatomy of a runaway agent's night
When an agent's bill jumps, the cause is often a loop, a recursive delegation, an injected instruction or a stolen key rather than growth. One illustrative overnight runaway, step by step, mapped to OWASP and MITRE, then the per-agent controls that would have stopped it.
Intertrace Threat Research
Technical paper10 min read
What to record for every agent action: a security telemetry schema
Agent logs inherited from performance monitoring say how long and how much, not who, on whose authority, or why it was allowed. A five-band event schema mapped to OpenTelemetry GenAI conventions, with privacy rules and retention tiers.
Intertrace
Guide10 min read
Threat model an AI agent in one afternoon
Most agent threat models study the model and miss the credentials. This five-pass, four-hour method models authority instead: data flow, real permissions, consequence tiers, everyone who writes into context and failure tracing, ending in one page with named owners.
Intertrace
+Stay current
New research, the week it publishes.
Subscribe with RSS, or email support@intertrace.ai with a topic you want covered. To see the controls in action, try the Simulation Lab or read the documentation.