From e66b897c232633cf0adc57ae19d78744162be716 Mon Sep 17 00:00:00 2001 From: Matteo Meucci Date: Thu, 13 Nov 2025 20:27:50 +0100 Subject: [PATCH] Update AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md --- ...ITG-APP-06_Testing_for_Agentic_Behavior_Limits.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/Document/content/tests/AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md b/Document/content/tests/AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md index 1b571e2..8a4f501 100644 --- a/Document/content/tests/AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md +++ b/Document/content/tests/AITG-APP-06_Testing_for_Agentic_Behavior_Limits.md @@ -1,4 +1,4 @@ -## AITG-APP-06 – Testing for Agentic Behavior Limits +# AITG-APP-06 – Testing for Agentic Behavior Limits ### Summary Agentic behavior limits refer to the safeguards placed around AI agents to prevent unintended autonomous actions. AI agents capable of planning and acting (e.g., Auto-GPT) may exceed user intent by generating sub-goals, refusing to halt, or misusing tools. This test verifies whether AI agents operate within their designed autonomy, respect user instructions (e.g., termination), and avoid unsafe or emergent behaviors like deception, recursive planning, or overreach. These tests are crucial to prevent misuse, ensure safety, and align agents with ethical and functional constraints. @@ -122,8 +122,8 @@ In 2023, GPT-4, when tested by ARC, hired a human on TaskRabbit to solve a CAPTC - [https://www.star-history.com/blog/agentic-security](https://www.star-history.com/blog/agentic-security) ### References -- OWASP Top 10 for LLM – LLM06: Excessive Agency – https://genai.owasp.org -- ARC Test on GPT-4 deception – https://www.vice.com/en/article/bvmv7v/gpt-4-taskrabbit-openai -- ChaosGPT Case Study – https://www.vice.com/en/article/m7gz3n/chaosgpt -- Prompt Flow Integrity (PFI) – https://arxiv.org/abs/2503.15547 -- SafeAgentBench – https://arxiv.org/abs/2412.13178 +- OWASP Top 10 for LLM – LLM06: Excessive Agency – [Link](https://genai.owasp.org/llmrisk/llm06-sensitive-information-disclosure/?utm_source=chatgpt.com) +- ARC Test on GPT-4 deception – [Link](https://www.vice.com/en/article/bvmv7v/gpt-4-taskrabbit-openai) +- ChaosGPT Case Study – [Link](https://www.vice.com/en/article/m7gz3n/chaosgpt) +- Prompt Flow Integrity (PFI) – [Link](https://arxiv.org/abs/2503.15547) +- SafeAgentBench – [Link](https://arxiv.org/abs/2412.13178)