Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
2,982 detections
Filters
Last updated
All Time
Detection languages
1,486
513
403
239
167
Contributors
986
658
304
198
109
Categories
916
822
673
590
452
Platforms
39,932
7,263
6,535
3,865
3,643
2,982
Products / Services
1,229
703
443
309
200
MITRE Techniques
1,053
824
459
384
348
CVEs
17
16
16
10
10
IDS Classtypes
24
7
7
4
3
IDS Protocols
49
1
Detects suspicious AI inference API activity indicative of credential compromise or misuse. The rule identifies three main signals: rapid authentication failures (credential stuffing), anomalous geolocation/ASN usage for successful requests, and successful API calls occurring after a key rotation event.
Detects high-volume API requests to an AI inference service that explicitly request gradient, confidence vector, or raw logit information. Such patterns are indicative of model inversion or training data reconstruction attacks, where an adversary attempts to extract private training data by exploiting detailed output from the model.
Detects attempts by users to extract the underlying system prompt from an LLM via prompt injection techniques or identifies successful leaks where the model's response matches a pre-calculated system prompt hash. Exposing system prompts can lead to the discovery of proprietary business logic, safety guidelines, and facilitate further adversarial jailbreaking.
Detects instances where AI agents are activated through external communication channels like webhooks, email, or chat, and the provided payload includes potentially malicious imperative instructions (e.g., 'delete', 'execute', 'send credentials'). The detection logic excludes activity originating from trusted, allow-listed internal sources to reduce false positives.
Detects adversarial attempts to infer whether specific records were used in an AI model's training set by analyzing raw prediction confidence scores and identifying suspicious query patterns such as near-duplicate submissions or systematic sweeps.
Detects when a victim user invokes an AI agent tool that has been potentially poisoned. This is identified by either a change in the tool definition hash sourced from unverified or untrusted locations, or by an invocation that results in unexpected, undocumented, or unauthorized secondary tool calls.
This rule monitors the behavior of AI agent orchestrators to detect potentially malicious activities, specifically focusing on unauthorized tool usage, deep tool invocation chains, and potential data exfiltration. It triggers when an agent invokes a tool outside of its authorized scope, performs an excessively deep sequence of tool calls, or accesses sensitive data followed by an outbound connection to an external endpoint, which is consistent with the MITRE ATLAS AML.T0053 technique.
Detects attempts to extract or override the system instructions of an Large Language Model (LLM) via prompt injection techniques, such as explicitly asking for the system prompt or commanding the model to ignore previous instructions.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
Detects LLM jailbreak attempts using a 'crescendo' multi-turn strategy. The rule monitors for sequences where an initial prompt contains known jailbreak-intent keywords leading to an AI refusal, followed shortly by a compliant AI response to a rephrased request within the same conversation session.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
Detects LLM jailbreak attempts using a 'crescendo' multi-turn strategy. The rule monitors for sequences where an initial prompt contains known jailbreak-intent keywords leading to an AI refusal, followed shortly by a compliant AI response to a rephrased request within the same conversation session.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
This rule detects high volumes of near-identical, incrementally-varied queries directed at an AI model API. This behavior is indicative of a black-box adversarial optimization loop, where an attacker attempts to infer model vulnerabilities, bypass safety filters, or perform model evasion by observing how slight input variations change model responses.
Detects potential training data poisoning attempts by monitoring anomalies in ingestion volume, shifts in label distribution, and high-volume submissions from feedback channels. This rule uses statistical analysis to identify spikes in data input or abnormal distributions that deviate from established historical baselines.
This rule detects potentially malicious activity related to AI agent tool manipulation. It flags two primary scenarios: 1) The invocation of an AI tool shortly after an unmanaged or potentially unauthorized modification to its definition, and 2) Detection of prompt injection patterns within the output generated by an AI agent tool, which could indicate an attacker attempting to override the agent's instructions.
This rule detects potentially malicious activity related to AI agent tool manipulation. It flags two primary scenarios: 1) The invocation of an AI tool shortly after an unmanaged or potentially unauthorized modification to its definition, and 2) Detection of prompt injection patterns within the output generated by an AI agent tool, which could indicate an attacker attempting to override the agent's instructions.
This rule detects anomalous behavior in AI agent tool invocations, specifically monitoring for high volumes of tool calls, high costs associated with token consumption, or suspicious tool interactions following a flagged indirect prompt injection. This behavior is indicative of potential resource exhaustion (denial of service) or abusive agent behavior as defined by the MITRE ATLAS framework for AI security.
Detects potential sensitive data leakage from an AI model response (AIGatewayLogs_CL). The rule monitors AI responses for patterns indicative of PII (SSN, credit card numbers), secrets (API keys, tokens), or sensitive document markers. It correlates these signals with RAG (Retrieval-Augmented Generation) ingestion logs to identify data cross-tenant violations or improper access to confidential/restricted datasets.
Page 31 of 142