---
name: tai-theme-frontier-safety-containment-and-deception
description: 'Use the Testing AI theme Frontier Safety, Containment, and Deception to plan, review, or teach related AI quality work. Applies concepts and techniques from the book to testing AI, AI-generated software, and non-deterministic systems when relevant.'
---

# Frontier Safety, Containment, and Deception

Skill name: `tai-theme-frontier-safety-containment-and-deception`

Based on **Testing AI: Engineering Confidence in AI Systems** by **Jason Arbon**.

## Theme Purpose

Use these approaches when testing hazardous capabilities, containment, manipulation, deception, evaluation awareness, and systems that may exceed human evaluator capability.

Apply these concepts when testing AI, AI-generated software, model-backed features, agents, search, chatbots, RAG systems, generated code, dynamic interfaces, or other software whose behavior can vary across runs, users, data, tools, or time.

## How To Use This Theme

- Identify the behavior, capability, risk, or release decision being evaluated.
- Choose the relevant concepts below and turn them into concrete eval cases, samples, traces, checks, rubrics, metrics, or release gates.
- Prefer evidence that supports a decision: ship, canary, hold, rollback, or collect more samples.
- Report by slices and severe failures when averages hide risk.
- Preserve enough evidence that another person or agent can understand what was tested, how it was measured, and why the recommendation follows.

## Concepts And Techniques To Apply

- Test hazardous capabilities, containment, sandbox escape, manipulation, persuasion, deception, scheming, and evaluation awareness.
- Use concrete safety benchmarks and red-team suites where appropriate, but inspect their scope and limitations.
- Test whether systems can fail safely, remain contained, respect capability boundaries, and avoid hidden goal pursuit.
- Assume future evaluators may be testing systems more capable than themselves; use layered, independent, adversarial evidence.

## Reporting Guidance

- State what was tested and what population the evidence represents.
- Explain uncertainty, missing coverage, severe failures, and known blind spots.
- Connect findings to a concrete decision or next action.
- Use topic-specific chapter skills only when deeper detail is needed; this theme skill should stand alone as practical guidance.