Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
29 detections
Filters
Last updated
All Time
Detection languages
9
8
7
5
Contributors
14
9
4
1
1
Categories
16
11
10
6
4
Platforms
16
12
3
2
1
Products / Services
15
11
4
3
1
MITRE Techniques
18,116
15,638
12,585
8,395
6,225
Detects the presence of known adversarial LLM jailbreak or prompt injection patterns embedded within source code comments. These patterns are designed to bypass safety filters or security controls of AI models when the source code is processed or analyzed by such systems.
Detects the presence of zero-width, non-printing, or bidirectional-override Unicode characters in command lines, commonly used for obfuscation or prompt injection.
Detects the presence of zero-width, non-printing, or bidirectional-override Unicode characters in command lines, commonly used for obfuscation or prompt injection.
Detects invisible/non-printing Unicode codepoints (zero-width chars, bidi overrides, variation selectors, Unicode tag characters, and Private Use Area icon-font codepoints) embedded in process command lines - covering both classic command/script obfuscation and the emerging use of hidden codepoints to smuggle AI prompt-injection payloads into content later processed by AI copilots/agents. Tightened to require 2+ high-confidence codepoint hits (or 1 from a known scripting host/LOLBin), and gates Private-Use-Area-only matches (a common legitimate icon-font false-positive source) to scripting hosts with 5+ hits, maps to T1027.018.
This rule detects potential prompt injection attacks against internal AI-powered assistants, agents, or LLM gateways. It monitors network HTTP requests for common patterns used to override system instructions, bypass safety filters, or force the model to adopt a privileged, unauthorized persona (e.g., administrator or debug mode). These techniques are often used in social engineering to manipulate the AI's output or security posture.
Detects the presence of hidden DOM elements containing suspicious AI instructional phrases (e.g., 'ignore previous instructions', 'you are now') served in HTML responses from known AI assistant platforms. This behavior is indicative of a Prompt-Forcing attack, where an adversary attempts to manipulate the AI model's behavior by injecting hidden instructions that the user cannot see but the AI consumes.
This search surfaces AWS Bedrock Claude prompts containing phrases commonly associated with prompt injection or jailbreak attempts, such as instruction overrides, persona switching, or requests to ignore prior guidance. These strings are not inherently malicious on their own - many of them (for example "act as", "you must", or "new persona") appear routinely in legitimate system prompts, few-shot examples, and role-play or creative-writing use cases. This is a hunting search rather than an anomaly detection because the base rate of benign matches is high; use it to pivot into the surrounding conversation and look for corroborating signals such as the phrase appearing mid-conversation rather than in an expected system/few-shot prompt, the user attempting to override guardrails or safety instructions rather than set a benign persona, instructions that follow the injected phrase requesting privilege escalation, data exfiltration, or access to restricted tools/resources, repeated or iterative attempts from the same user/role after being refused, and injected phrases originating from untrusted input (e.g., retrieved documents or tool output) rather than the human user turn.
Detects instances where Zscaler AI Protect (AI Guard) flags attempts to perform prompt injection or jailbreak against Large Language Model (LLM) or GenAI applications. The rule monitors blocked requests and allowed requests that exceed a high confidence score threshold (0.75), identifying potential malicious interactions with generative AI services.
Page 2 of 2