MITRE ATLAS — Adversarial Threat Landscape for AI Systems
AI attack techniques, explained for teams building with agents
- 91Reconnaissance: 9 techniques, 0 covered by a Securitain guide
- 112Resource Development: 11 techniques, 1 covered by a Securitain guide
- 113AI Attack Adaptation: 11 techniques, 1 covered by a Securitain guide
- 104Initial Access: 10 techniques, 4 covered by a Securitain guide
- 45AI Model Access: 4 techniques, 1 covered by a Securitain guide
- 66Execution: 6 techniques, 3 covered by a Securitain guide
- 137Persistence: 13 techniques, 6 covered by a Securitain guide
- 48Privilege Escalation: 4 techniques, 4 covered by a Securitain guide
- 219Defense Evasion: 21 techniques, 2 covered by a Securitain guide
- 710Credential Access: 7 techniques, 4 covered by a Securitain guide
- 1111Discovery: 11 techniques, 1 covered by a Securitain guide
- 612Lateral Movement: 6 techniques, 3 covered by a Securitain guide
- 913Collection: 9 techniques, 1 covered by a Securitain guide
- 514Command and Control: 5 techniques, 1 covered by a Securitain guide
- 515Exfiltration: 5 techniques, 3 covered by a Securitain guide
- 1016Impact: 10 techniques, 1 covered by a Securitain guide
ATLAS is a separate framework from ATT&CK
ATLAS techniques use AML.* identifiers and are never treated as ATT&CK technique IDs. Where ATLAS itself lists an ATT&CK reference we show it as such; any other relationship on this site is an explicit, reviewed Securitain judgement. See also the MITRE ATT&CK tactic guides.Securitain AI security guides
Start with the threats that matter for AI agents
- 5 ATLAS techniques
Direct and indirect prompt injection
What is prompt injection, and what is the difference between direct and indirect? - 3 ATLAS techniques
AI agent tool misuse and unauthorized actions
How can an AI agent be manipulated into taking actions nobody approved? - 6 ATLAS techniques
Agent identity, permissions and credential exposure
How should AI agent permissions and credentials be managed in AWS? - 4 ATLAS techniques
RAG, memory and retrieved-content poisoning
How can retrieval-augmented generation (RAG) be poisoned? - 4 ATLAS techniques
Sensitive information disclosure and AI-enabled exfiltration
Can an AI agent expose sensitive data? - 6 ATLAS techniques
Model, dataset and agent-tool supply chain
What are AI supply-chain threats? - 4 ATLAS techniques
Model and AI service access management
How should access to AI models and services like Amazon Bedrock be controlled? - 4 ATLAS techniques
Cross-system agent lateral movement and execution boundaries
How can an AI agent become a bridge between systems for an attacker?
ATLAS matrix order
ATLAS tactics
- 01 · AML.TA0002Reconnaissance9 techniques
- 02 · AML.TA0003Resource Development11 techniques
- 03 · AML.TA0001AI Attack Adaptation11 techniques
- 04 · AML.TA0004Initial Access10 techniques
- 05 · AML.TA0000AI Model Access4 techniques
- 06 · AML.TA0005Execution6 techniques
- 07 · AML.TA0006Persistence13 techniques
- 08 · AML.TA0012Privilege Escalation4 techniques
- 09 · AML.TA0007Defense Evasion21 techniques
- 10 · AML.TA0013Credential Access7 techniques
- 11 · AML.TA0008Discovery11 techniques
- 12 · AML.TA0015Lateral Movement6 techniques
- 13 · AML.TA0009Collection9 techniques
- 14 · AML.TA0014Command and Control5 techniques
- 15 · AML.TA0010Exfiltration5 techniques
- 16 · AML.TA0011Impact10 techniques
Searchable
Technique explorer
129 techniques shown
- AML.T0000
Search Open Technical Databases
Reconnaissance · Maturity: Realized · ATLAS-listed ATT&CK reference: T1596
Adversaries may search for publicly available research and technical documentation to learn how and where AI is used within a victim organization. The adversary can use this information to identify targets for attack,…
- AML.T0001
Search Open AI Vulnerability Analysis
Reconnaissance · Maturity: Demonstrated
Much like the Search Open Technical Databases, there is often ample research available on the vulnerabilities of common AI models. Once a target has been identified, an adversary will likely try to identify any pre-ex…
- AML.T0002
Acquire Public AI Artifacts
Resource Development · Maturity: Realized
Adversaries may search public sources, including cloud storage, public-facing services, and software or data repositories, to identify AI artifacts. These AI artifacts may include the software stack used to train and…
- AML.T0003
Search Victim-Owned Websites
Reconnaissance · Maturity: Demonstrated · ATLAS-listed ATT&CK reference: T1594
Adversaries may search websites owned by the victim for information that can be used during targeting. Victim-owned websites may contain technical details about their AI-enabled products or services. Victim-owned webs…
- AML.T0004
Search Application Repositories
Reconnaissance · Maturity: Demonstrated
Adversaries may search open application repositories during targeting. Examples of these include Google Play, the iOS App store, the macOS App Store, and the Microsoft Store. Adversaries may craft search queries seek…
- AML.T0005
Replicate AI Model Behavior
Resource Development · AI Attack Adaptation · Maturity: Realized
Adversaries may replicate an AI model's behavior or selected capabilities by training or adapting another model to approximate the target's responses, decision regions, or other characteristics. The result may be a su…
- AML.T0006
Active Scanning
Reconnaissance · Maturity: Realized · ATLAS-listed ATT&CK reference: T1595
Adversaries may actively scan for publicly reachable AI systems and resources for targeting. They may identify resources or systems that are misconfigured or are known vulnerable versions. This is distinct from other…
- AML.T0007
Discover AI Artifacts
Discovery · Maturity: Demonstrated
Adversaries may search private sources to identify AI learning artifacts that exist on the system and gather information about them. These artifacts can include the software stack used to train and deploy models, trai…
- AML.T0008
Acquire Infrastructure
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1583
Adversaries may buy, lease, or rent infrastructure for use throughout their operation. A wide variety of infrastructure exists for hosting and orchestrating adversary operations. Infrastructure solutions include physi…
- AML.T0010
Supply Chain Compromise
Initial Access · Maturity: Realized · ATLAS-listed ATT&CK reference: T1195
Adversaries may manipulate products, components, or their development and delivery mechanisms before they reach downstream consumers to compromise data or systems. Manipulation may involve source code, dependencies, d…
- AML.T0011
User Execution
Execution · Maturity: Realized
An adversary may rely upon a user to load, execute, interpret, or otherwise use a malicious or unsafe artifact. In AI systems, user execution may include loading a model or causing an inference runtime to process exec…
- AML.T0012
Valid Accounts
Initial Access · Privilege Escalation · Lateral Movement · Maturity: Realized · ATLAS-listed ATT&CK reference: T1078
Adversaries may obtain and abuse credentials of existing accounts as a means of gaining Initial Access. Credentials may take the form of usernames and passwords of individual user accounts or API keys that provide acc…
- AML.T0013
Discover AI Model Ontology
Discovery · Maturity: Demonstrated
Adversaries may discover the ontology of an AI model's output space, for example, the types of objects a model can detect. The adversary may discover the ontology by repeated queries to the model, forcing it to enumer…
- AML.T0014
Discover AI Model Family
Discovery · Maturity: Feasible
Adversaries may discover the general family of a model. General information about the model may be revealed in documentation, or the adversary may use carefully constructed examples and analyze the model's responses t…
- AML.T0015
Evade AI Model
Initial Access · Defense Evasion · Impact · Maturity: Realized
Adversaries can Craft Adversarial Data that prevents an AI model from correctly identifying the contents of the data or Generate Deepfakes that fools an AI model expecting authentic data. This technique can be used t…
- AML.T0016
Obtain Capabilities
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1588
Adversaries may search for and obtain software capabilities for use in their operations. Capabilities may be specific to AI-based attacks Adversarial AI Attack Implementations or generic software tools repurposed for…
- AML.T0017
Develop Capabilities
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1587
Adversaries may develop their own capabilities to support operations. This process encompasses identifying requirements, building or adapting solutions, validating or packaging capabilities, and preparing them for dep…
- AML.T0018
Manipulate AI Model
AI Attack Adaptation · Persistence · Maturity: Realized
Adversaries may manipulate an AI model artifact or its bundled components to change AI system behavior, introduce malicious code, or establish persistent malicious functionality. This may include modifying model weigh…
- AML.T0020
Training Data Poisoning
Persistence · Maturity: Realized
Adversaries may manipulate data used for training or fine-tuning an AI model to influence the resulting model's behavior. Adversaries may add, remove, or modify data samples; alter labels or annotations; or manipulate…
- AML.T0021
Establish Accounts
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1585
Adversaries may create accounts with various services for use in targeting, to gain access to resources needed to further their operations, or for victim impersonation.
- AML.T0024
Exfiltration via AI Inference API
Exfiltration · Maturity: Realized
Adversaries may exfiltrate information from a victim-controlled AI model or service through responses returned by its inference API. Transferred information may include predictions, scores, embeddings, generated conte…
- AML.T0025
Exfiltration via Cyber Means
Exfiltration · Maturity: Realized
Adversaries may exfiltrate AI artifacts or other information relevant to their goals via traditional cyber means. See the ATT&CK Exfiltration tactic for more information.
- AML.T0029
Denial of AI Service
Impact · Maturity: Demonstrated
Adversaries may target AI-enabled systems with a flood of requests for the purpose of degrading or shutting down the service. Since many AI systems require significant amounts of specialized compute, they are often ex…
- AML.T0031
Erode AI Model Integrity
Impact · Maturity: Realized
Adversaries may degrade the target model's performance with adversarial data inputs to erode confidence in the system over time. This can lead to the victim organization wasting time and money both attempting to fix t…
- AML.T0034
Cost Harvesting
Impact · Maturity: Feasible
Adversaries may deliberately drive a victim's AI services beyond normal operating capacity with the intent of increasing the cost of services. This may be achieved via high-volume, low-complexity queries (Excessive Qu…
- AML.T0035
AI Artifact Collection
Collection · Maturity: Realized
Adversaries may collect AI artifacts for Exfiltration or for use in AI Attack Adaptation. AI artifacts include models and datasets as well as other telemetry data produced when interacting with a model.
- AML.T0036
Data from Information Repositories
Collection · Maturity: Realized · ATLAS-listed ATT&CK reference: T1213
Adversaries may leverage information repositories to mine valuable information. Information repositories are tools that allow for storage of information, typically to facilitate collaboration or information sharing be…
- AML.T0037
Data from Local System
Collection · Maturity: Realized · ATLAS-listed ATT&CK reference: T1005
Adversaries may search local system sources, such as file systems and configuration files or local databases, to find files of interest and sensitive data prior to Exfiltration. This can include basic fingerprinting…
- AML.T0040
AI Model Inference API Access
AI Model Access · Maturity: Realized
Adversaries may gain access to a model via legitimate access to the inference API. Inference API access can be a source of information to the adversary (Discover AI Model Ontology, Discover AI Model Family), a means o…
- AML.T0041
Physical Environment Access
AI Model Access · Maturity: Demonstrated
In addition to the attacks that take place purely in the digital domain, adversaries may also exploit the physical environment for their attacks. If the model is interacting with data collected from the real world in…
- AML.T0042
Verify Attack
AI Attack Adaptation · Maturity: Realized
Adversaries can verify the efficacy of their attack via an inference API or access to an offline copy of the target model. This gives the adversary confidence that their approach works and allows them to carry out the…
- AML.T0043
Craft Adversarial Data
AI Attack Adaptation · Maturity: Realized
Adversarial data are inputs to an AI model that have been modified such that they cause the adversary's desired effect in the target model. Effects can range from misclassification, to missed detections, to maximizing…
- AML.T0044
Full AI Model Access
AI Model Access · Maturity: Demonstrated
Adversaries may gain full "white-box" access to an AI model. This means the adversary has complete knowledge of the model architecture, its parameters, and class ontology. They may exfiltrate the model to Craft Advers…
- AML.T0046
Spamming AI System with Chaff Data
Impact · Maturity: Feasible
Adversaries may spam the AI system with chaff data that causes increase in the number of detections. This can cause analysts at the victim organization to waste time reviewing and correcting incorrect inferences. Adv…
- AML.T0047
AI-Enabled Product or Service
AI Model Access · Maturity: Realized
Adversaries may use a product or service that uses artificial intelligence under the hood to gain access to the underlying AI model. This type of indirect model access may reveal details of the AI model or its inferen…
- AML.T0048
External Harms
Impact · Maturity: Realized
Adversaries may abuse their access to a victim system and use its resources or capabilities to further their goals by causing harms external to that system. These harms could affect the organization (e.g. Financial Ha…
- AML.T0049
Exploit Public-Facing Application
Initial Access · Maturity: Realized · ATLAS-listed ATT&CK reference: T1190
Adversaries may attempt to take advantage of a weakness in an Internet-facing computer or program using software, data, or commands in order to cause unintended or unanticipated behavior. The weakness in the system ca…
- AML.T0050
Command and Scripting Interpreter
Execution · Maturity: Realized · ATLAS-listed ATT&CK reference: T1059
Adversaries may abuse command and script interpreters to execute commands, scripts, or binaries. These interfaces and languages provide ways of interacting with computer systems and are a common feature across many di…
- AML.T0051
LLM Prompt Injection
Execution · Maturity: Realized
An adversary may craft malicious prompts as inputs to an LLM that cause the LLM to act in unintended ways. These "prompt injections" are often designed to cause the model to ignore aspects of its original instructions…
- AML.T0052
Phishing
Initial Access · Lateral Movement · Maturity: Realized · ATLAS-listed ATT&CK reference: T1566
Adversaries may send phishing messages to gain access to victim systems. All forms of phishing are electronically delivered social engineering. Phishing can be targeted, known as spearphishing. In spearphishing, a spe…
- AML.T0053
AI Agent Tool Invocation
Execution · Privilege Escalation · Lateral Movement · Maturity: Demonstrated
Adversaries may use their access to an AI agent to invoke tools the agent has access to. LLMs are often connected to other services or resources via tools to increase their capabilities. Tools may include integrations…
- AML.T0054
LLM Jailbreak
Privilege Escalation · Defense Evasion · Maturity: Realized
Adversaries may induce a large language model (LLM) to ignore, circumvent, or override its safety/alignment behaviors and/or guardrails to elicit outputs the model is intended to withhold. Once jailbroken, the LLM may…
- AML.T0055
Unsecured Credentials
Credential Access · Maturity: Realized · ATLAS-listed ATT&CK reference: T1552
Adversaries may search compromised systems to find and obtain insecurely stored credentials. These credentials can be stored and/or misplaced in many locations on a system, including plaintext files (e.g. bash history…
- AML.T0059
Erode Dataset Integrity
Impact · Maturity: Demonstrated
Adversaries may poison or manipulate portions of a dataset to reduce its usefulness, reduce trust, and cause users to waste resources correcting errors.
- AML.T0060
Publish Hallucinated Entities
Resource Development · Maturity: Demonstrated
Adversaries may create an entity they control, such as a software package, website, or email address corresponding to a source hallucinated by an LLM. The hallucinations may take the form of package names, commands, U…
- AML.T0061
LLM Prompt Self-Replication
Persistence · Maturity: Demonstrated
Adversaries may introduce a prompt designed to replicate itself in a large language model's (LLM) output. The reproduced instructions may appear in generated messages, documents, source-code, files, or other artifacts…
- AML.T0062
Discover LLM Hallucinations
Discovery · Maturity: Demonstrated
Adversaries may prompt large language models and identify hallucinated entities. They may request software packages, commands, URLs, organization names, or e-mail addresses, and identify hallucinations with no connect…
- AML.T0063
Discover AI Model Outputs
Discovery · Maturity: Realized
Adversaries may discover what information an AI model or service exposes through its outputs. They may investigate both the intended responses and additional model-generated information exposed through the surrounding…
- AML.T0064
Gather RAG-Indexed Targets
Reconnaissance · Maturity: Demonstrated
Adversaries may identify data sources used in retrieval augmented generation (RAG) systems for targeting purposes. By pinpointing these sources, attackers can focus on poisoning or otherwise manipulating the external…
- AML.T0065
LLM Prompt Crafting
AI Attack Adaptation · Maturity: Realized
Adversaries may use their acquired knowledge of the target generative AI system to craft prompts that bypass its defenses and allow malicious instructions to be executed. They may iterate on prompts to ensure that the…
- AML.T0066
Retrieval Content Crafting
AI Attack Adaptation · Maturity: Demonstrated
Adversaries may write content designed to be retrieved by user queries and influence a user of the system in some way. This abuses the trust the user has in the system. The crafted content can be combined with a prom…
- AML.T0067
LLM Trusted Output Components Manipulation
Defense Evasion · Maturity: Demonstrated
Adversaries may utilize prompts to a large language model (LLM) which manipulate various components of its response in order to make it appear trustworthy to the user. This helps the adversary continue to operate in t…
- AML.T0068
LLM Prompt Obfuscation
Defense Evasion · Maturity: Realized
Adversaries may hide or otherwise obfuscate prompt injections or retrieval content to avoid detection from humans, large language model (LLM) guardrails, or other detection mechanisms. Text inputs may be obfuscated b…
- AML.T0069
Discover LLM System Information
Discovery · Maturity: Demonstrated
The adversary is trying to discover something about the large language model's (LLM) system information. This may be found in a configuration file containing the system instructions or extracted via interactions with…
- AML.T0070
RAG Poisoning
Persistence · Maturity: Demonstrated
Adversaries may inject malicious content into data indexed by a retrieval augmented generation (RAG) system to contaminate a future thread through RAG-based search results. This may be accomplished by placing manipula…
- AML.T0071
False RAG Entry Injection
Defense Evasion · Maturity: Demonstrated
Adversaries may introduce false entries into a victim's retrieval augmented generation (RAG) database. Content designed to be interpreted as a document by the large language model (LLM) used in the RAG system is inclu…
- AML.T0072
Cyber Communication Channel
Command and Control · Maturity: Realized
Adversaries may establish or use cyber communication channels for command and control. A channel may use any network protocol, service, repository, relay, or other intermediary and may operate synchronously or asynchr…
- AML.T0073
Impersonation
Defense Evasion · Maturity: Realized · ATLAS-listed ATT&CK reference: T1684.001
Adversaries may impersonate an existing person or organization, trusted AI agent, or present a fabricated persona as a trustworthy or independent party, to persuade a target to act on their behalf. Claimed identities…
- AML.T0074
Masquerading
Defense Evasion · Maturity: Realized · ATLAS-listed ATT&CK reference: T1036
Adversaries may attempt to manipulate features of their artifacts to make them appear legitimate or benign to users and/or security tools. Masquerading occurs when the name or location of an object, legitimate or mali…
- AML.T0075
Enterprise Resource Discovery
Discovery · Maturity: Realized
Adversaries may discover resources available within an enterprise environment and information needed to identify or access them. Resources may include accounts, groups, systems, devices, virtual machines, files, direc…
- AML.T0076
Corrupt AI Model
Defense Evasion · Maturity: Realized
An adversary may purposefully corrupt a malicious AI model file so that it cannot be successfully deserialized in order to evade detection by a model scanner. The corrupt model may still successfully execute malicious…
- AML.T0077
LLM Response Rendering
Exfiltration · Maturity: Demonstrated
Adversaries may induce a large language model (LLM) to respond with private information structured in a reference to external content that, when rendered by the user's client, makes a request to an adversary-controlle…
- AML.T0078
Drive-by Compromise
Initial Access · Maturity: Demonstrated · ATLAS-listed ATT&CK reference: T1189
Adversaries may gain access to an AI system through a user visiting a website over the normal course of browsing, or an AI agent retrieving information from the web on behalf of a user. Websites can contain an LLM Pro…
- AML.T0079
Stage Capabilities
Resource Development · Maturity: Realized
Adversaries may upload, install, or otherwise set up capabilities that can be used during targeting. To support their operations, an adversary may need to take capabilities they developed (Develop Capabilities) or obt…
- AML.T0080
AI Agent Context Poisoning
Persistence · Maturity: Realized
Adversaries may attempt to manipulate the context used by an AI agent's large language model (LLM) to influence the responses it generates or actions it takes. This allows an adversary to persistently change the behav…
- AML.T0081
Modify AI Agent Configuration
Persistence · Defense Evasion · Maturity: Demonstrated
Adversaries may modify the configuration files for AI agents on a system. This allows malicious changes to persist beyond the life of a single agent and affects any agents that share the configuration. Configuration…
- AML.T0082
RAG Credential Harvesting
Credential Access · Maturity: Demonstrated
Adversaries may attempt to use their access to a large language model (LLM) on the victim's system to collect credentials. Credentials may be stored in internal documents which can inadvertently be ingested into a RAG…
- AML.T0083
Credentials from AI Agent Configuration
Credential Access · Maturity: Demonstrated
Adversaries may access the credentials of other tools or services on a system from the configuration of an AI agent.
- AML.T0084
Discover AI Agent Configuration
Discovery · Maturity: Demonstrated
Adversaries may attempt to discover configuration information for AI agents present on the victim's system. Agent configurations can include tools or services they have access to. Adversaries may directly access agen…
- AML.T0085
Data from AI Services
Collection · Maturity: Demonstrated
Adversaries may use their access to victim AI-enabled services to collect proprietary or otherwise sensitive information available to the system. As organizations adopt generative and agentic AI in centralized service…
- AML.T0086
Exfiltration via AI Agent Tool Invocation
Exfiltration · Maturity: Realized
AI agent tools capable of performing write operations may be invoked to exfiltrate data to an adversary. Sensitive information can be encoded into the tool's input parameters and transmitted to an adversary-controlled…
- AML.T0087
Gather Victim Identity Information
Reconnaissance · Maturity: Realized · ATLAS-listed ATT&CK reference: T1589
Adversaries may gather information about the victim's identity that can be used during targeting. Information about identities may include a variety of details, including personal data (ex: employee names, email addre…
- AML.T0088
Generate Deepfakes
AI Attack Adaptation · Maturity: Realized
Adversaries may use generative artificial intelligence (GenAI) to create synthetic media (i.e. imagery, video, audio, and text) that appear authentic. These "deepfakes" may mimic a real person or depict fictional pers…
- AML.T0089
Enterprise Environment Discovery
Discovery · Maturity: Realized
Adversaries may discover configurations, conditions, relationships, activity, and other characteristics of an enterprise environment. This may include system and network configuration; active connections, processes, w…
- AML.T0090
OS Credential Dumping
Credential Access · Maturity: Demonstrated · ATLAS-listed ATT&CK reference: T1003
Adversaries may extract credentials from OS caches, application memory, or other sources on a compromised system. Credentials are often in the form of a hash or clear text, and can include usernames and passwords, app…
- AML.T0091
Use Alternate Authentication Material
Lateral Movement · Maturity: Realized · ATLAS-listed ATT&CK reference: T1550
Adversaries may use alternate authentication material, such as password hashes, Kerberos tickets, and application access tokens, in order to move laterally within an environment and bypass normal system access control…
- AML.T0092
Manipulate User LLM Chat History
Defense Evasion · Maturity: Demonstrated
Adversaries may manipulate a user's large language model (LLM) chat history to cover the tracks of their malicious behavior. They may hide persistent changes they have made to the LLM's behavior, or obscure their atte…
- AML.T0093
Prompt Infiltration via Public-Facing Application
Initial Access · Persistence · Maturity: Realized
An adversary may introduce malicious prompts into the victim's system via a public-facing application with the intention of it being ingested by an AI at some point in the future and ultimately having a downstream eff…
- AML.T0094
Delay Execution of LLM Instructions
Defense Evasion · Maturity: Demonstrated
Adversaries may include instructions to be followed by the AI system in response to a future event, such as a specific keyword or the next interaction, in order to evade detection or bypass controls placed on the AI s…
- AML.T0095
Search Open Websites/Domains
Reconnaissance · Maturity: Realized · ATLAS-listed ATT&CK reference: T1593
Adversaries may search public websites and/or domains for information about victims and AI targets that can be used during targeting and to discover opportunities for successful attacks. Information about a victim's a…
- AML.T0096
AI Service API
Command and Control · Maturity: Realized
Adversaries may communicate using the API of an AI service on the victim's system. The adversary's commands to the victim system, and often the results, are embedded in the normal traffic of the AI service. An AI ser…
- AML.T0097
Virtualization/Sandbox Evasion
Defense Evasion · Maturity: Realized · ATLAS-listed ATT&CK reference: T1497
Adversaries may employ various means to detect and avoid virtualization and analysis environments. This may include changing behaviors based on the results of checks for the presence of artifacts indicative of a virtu…
- AML.T0098
AI Agent Tool Credential Harvesting
Credential Access · Maturity: Demonstrated
Adversaries may attempt to use their access to an AI agent on the victim's system to retrieve data from available agent tools to collect credentials. Agent tools may connect to a wide range of sources that may contain…
- AML.T0099
AI Agent Tool Data Poisoning
Persistence · Maturity: Feasible
Adversaries may manipulate data in a victim-controlled, trusted, or connected data source that is accessible through an AI agent tool. Poisoned data may be placed in document repositories, databases, source-code repos…
- AML.T0100
AI Agent Clickbait
Execution · Maturity: Demonstrated
Adversaries may craft deceptive content designed to bait computer-using AI agents or AI web browsers and tools into taking unintended actions, such as clicking buttons, copying code, or navigating to specific web page…
- AML.T0101
Data Destruction via AI Agent Tool Invocation
Impact · Maturity: Realized
Adversaries may invoke an AI agent's tool capable of performing mutative operations to perform Data Destruction. Adversaries may destroy data and files on specific systems or in large numbers on a network to interrupt…
- AML.T0102
Generate Malicious Commands
AI Attack Adaptation · Maturity: Realized
Adversaries may use large language models (LLMs) to dynamically generate malicious commands from natural language. Dynamically generated commands may be harder to detect as the attack signature is constantly changing.…
- AML.T0103
Deploy AI Agent
Execution · Maturity: Realized
Adversaries may launch AI agents in the victim's environment to execute actions on their behalf. AI agents may have access to a wide range of tools and data sources, as well as permissions to access and interact with…
- AML.T0105
Escape to Host
Privilege Escalation · Maturity: Realized · ATLAS-listed ATT&CK reference: T1611
Adversaries may break out of a container or virtualized environment to gain access to the underlying host. This can allow an adversary access to other containerized or virtualized resources from the host level or to t…
- AML.T0106
Exploitation for Credential Access
Credential Access · Maturity: Realized · ATLAS-listed ATT&CK reference: T1212
Adversaries may exploit software vulnerabilities in an attempt to collect credentials. Exploitation of a software vulnerability occurs when an adversary takes advantage of a programming error in a program, service, or…
- AML.T0107
Exploitation for Defense Evasion
Defense Evasion · Maturity: Demonstrated · ATLAS-listed ATT&CK reference: T1211
Adversaries may exploit a system or application vulnerability to bypass security features. Exploitation of a vulnerability occurs when an adversary takes advantage of a programming error in a program, service, or with…
- AML.T0108
AI Agent
Command and Control · Maturity: Demonstrated
Adversaries may abuse AI agents present on the victim's system for command and control. AI agents are often granted access to tools that can execute shell commands, reach out to the internet, and interact with other s…
- AML.T0109
AI Supply Chain Rug Pull
Defense Evasion · Maturity: Realized
Adversaries may publish legitimate AI components or software, gain user adoption, then push an update with a malicious variant, leading to Supply Chain Compromise. More scrutiny is often placed on a supply chain depen…
- AML.T0110
AI Agent Tool Poisoning
Persistence · Maturity: Realized
Adversaries may poison tools used by AI agents by introducing or modifying malicious content or behavior in a tool's model-visible definition, executable implementation, or runtime responses. Tools may include built-i…
- AML.T0111
AI Supply Chain Reputation Inflation
Defense Evasion · Maturity: Demonstrated
AI Supply Chain Reputation Inflation is the process of building or leveraging genuinely credible-looking trust signals to increase the perceived legitimacy of AI supply chain components, with the goal of driving adopt…
- AML.T0112
Machine Compromise
Impact · Maturity: Demonstrated
Adversaries may compromise a machine by exploiting or manipulating AI-enabled components on the system. Compromising a victim system allows the adversary to execute arbitrary code, steal credentials, exfiltrate data,…
- AML.T0113
Steal Web Session Cookie
Credential Access · Maturity: Demonstrated · ATLAS-listed ATT&CK reference: T1539
Adversaries may steal web application or service session cookies used to authenticate users to AI-enabled systems, AI services, or supporting enterprise applications. Web applications often use session cookies as aut…
- AML.T0114
AI Service Web Interface
Command and Control · Maturity: Demonstrated
Adversaries may communicate with compromised systems by abusing the web interface of an AI service as an intermediary command-and-control relay. Instead of connecting directly to attacker-controlled infrastructure or…
- AML.T0115
Publish Poisoned AI Artifacts
Resource Development · Maturity: Realized
Adversaries may create or modify AI artifacts and publish them through public or shared distribution channels to facilitate compromise of downstream AI systems. Poisoned AI artifacts may include datasets, models, and…
- AML.T0116
Autonomous Reconnaissance
Reconnaissance · Maturity: Realized
Adversaries may use autonomous AI agents to conduct Reconnaissance activities. Given an objective, target, or partial lead, an agent may autonomously determine what information to obtain and how to investigate it. It…
- AML.T0117
Autonomous Attack-Path Adaptation
AI Attack Adaptation · Maturity: Realized
Adversaries may use an AI agent to autonomously construct and repeatedly revise an attack path toward an adversary-defined objective. Given a high-level objective, the system may derive intermediate objectives, identi…
- AML.T0118
Autonomous AI Agent Communication
AI Attack Adaptation · Maturity: Realized
Autonomous AI agents may exchange operational information with other AI agents, sub-agents, or independent agent runs. Exchanged information may include discoveries, objectives, tasking, capabilities, credentials, con…
- AML.T0119
Exploit Automated Artifact Processing Pipeline
Initial Access · Maturity: Realized
Adversaries may submit, publish, or modify an artifact in a way that triggers an automated processing pipeline. As the pipeline handles the artifact, adversary-controlled content or configuration may exploit a weaknes…
- AML.T0120
AI Artifact Repository
Command and Control · Maturity: Realized
Adversaries may repurpose AI artifact repositories as asynchronous command-and-control channels. Commands, payloads, or tasking may be placed in repository objects for a compromised system to retrieve or poll. The com…
- AML.T0121
AI Agent Environment Reconstruction
Persistence · Maturity: Realized
An autonomous AI agent may reconstruct the environment needed to continue an existing operation after its execution environment is lost, reset, replaced, denied, or made unusable. The agent may recover or recreate too…
- AML.T0122
Exploitation of Remote Services
Lateral Movement · Maturity: Realized · ATLAS-listed ATT&CK reference: T1210
Adversaries may exploit a software or design weakness in a service reachable from their current environment to gain unauthorized access to another system, component, network, or trust boundary. Exploitation may allow…
- AML.T0123
Obfuscated Files or Information
Defense Evasion · Maturity: Realized · ATLAS-listed ATT&CK reference: T1027
Adversaries may attempt to make an executable or file difficult to discover or analyze by encrypting, encoding, or otherwise obfuscating its contents on the system or in transit. This is common behavior that can be us…
- AML.T0124
Autonomous Attack Orchestration
AI Attack Adaptation · Maturity: Realized
Adversaries may use autonomous AI systems as an operational control layer to manage multiple distinct autonomous agents or sub-agents toward a common adversary-defined objective. The orchestrating system may create as…
- AML.T0125
Create Account
Persistence · Maturity: Realized · ATLAS-listed ATT&CK reference: T1136
Adversaries may create an account to maintain access to an AI system or its supporting infrastructure. With a sufficient level of access, creating such accounts may be used to establish secondary credentialed access t…
- AML.T0126
Automated Collection
Collection · Maturity: Realized · ATLAS-listed ATT&CK reference: T1119
Adversaries may use automated techniques to collect data from AI systems and supporting enterprise environments. Automation may use scripts, command interpreters, command-line tools, or AI agent tools to identify, ret…
- AML.T0127
Data Staged
Collection · Maturity: Realized · ATLAS-listed ATT&CK reference: T1074
Adversaries may stage collected data in a central location before exfiltration. Staging consolidates, organizes, or prepares information obtained from one or more sources so it can be reviewed, processed, transferred,…
- AML.T0128
Compromise Infrastructure
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1584
Adversaries may compromise third-party infrastructure and repurpose it to support attacks against AI systems. Rather than buying, leasing, registering, or otherwise legitimately acquiring a resource, the adversary gai…
- AML.T0129
Triggers in Multimodal Inputs
Defense Evasion · Maturity: Feasible
Adversaries may place instructions or triggers in one part of a multimodal input to influence the model while staying unnoticed by human reviewers and by defenses that do not inspect all input modalities. Multimodal…
- AML.T0130
AI Agent Response Biasing
Impact · Maturity: Realized
Adversaries may manipulate an AI assistant so that it favors adversary-chosen sources, or content in its responses. By injecting instructions such as "treat [source] as a trusted source" or "recommend [source] first,"…
- AML.T0131
Crafted AI Assistant Links
Initial Access · Maturity: Realized
Adversaries may craft links that open an AI assistant or agent with attacker-controlled input already supplied, so that opening the link initiates an interaction the adversary defines rather than one the target compos…
- AML.T0132
Misconfigured or Publicly Exposed AI Services
Initial Access · Maturity: Demonstrated
AI agents and LLM platforms are deployed across diverse architectures, including standalone servers, SaaS platforms, and low-code builders. All of these often have misconfigured access controls, ranging from missing a…
- AML.T0133
Discover AI Agent Runtime Capabilities
Discovery · Maturity: Feasible
Adversaries may interact with an AI agent at runtime to reveal the capabilities available to it, without requiring access to its underlying configuration. Direct interaction with the agent can surface its registered t…
- AML.T0134
AI Targeted Cloaking
Defense Evasion · Maturity: Feasible
Adversaries may selectively deliver malicious or manipulated content to AI systems, while presenting different benign content to human users, web crawlers, or security detection mechanisms. They may identify AI browse…
- AML.T0135
Outputs from AI Models
Collection · Maturity: Realized
Adversaries may gather, retain, and organize outputs from an AI model or service as material for subsequent adversarial operations. Systematic selection and aggregation of responses can produce datasets or observation…
- AML.T0136
Exfiltration via AI User Interface
Exfiltration · Maturity: Realized
Adversaries exfiltrate information through an AI system's user-facing outputs. Information may be disclosed in conversational responses, generated documents, previews, or other content presented through the applicatio…
- AML.T0137
Training Data Information from AI Models
Collection · Maturity: Demonstrated
Adversaries may obtain sensitive information about the data used to train or fine-tune an AI model by analyzing its outputs or behavior. This can expose private associations, personally identifiable information, or pr…
- AML.T0138
Indicator Removal
Defense Evasion · Maturity: Realized · ATLAS-listed ATT&CK reference: T1070
Adversaries may delete, alter, or falsify records and artifacts to conceal their activity, impede investigation, or make unauthorized actions appear legitimate. Targeted artifacts may include logs, command histories,…
- AML.T0139
Compromise Accounts
Resource Development · Maturity: Realized · ATLAS-listed ATT&CK reference: T1586
Adversaries may compromise existing accounts with services that can support targeting or subsequent operations. These accounts may provide established identities, trusted relationships, access to resources, or funded…
- AML.T0140
Centralized AI Service Request Routing
Defense Evasion · Maturity: Realized
Adversaries may distribute coordinated AI service requests across multiple pathways to impede detection, conceal the coordinated nature of the requests, and evade AI provider enforcement. Distributing activity can lim…
- AML.T0141
AI Service Request Metadata Sanitization
Defense Evasion · Maturity: Realized
Adversaries may remove or replace identifying information in request metadata before submitting requests to an AI service to impede attribution and correlation of coordinated activity. Sanitization may be enforced aut…
- AML.T0142
Event Triggered Execution
Persistence · Maturity: Realized · ATLAS-listed ATT&CK reference: T1546
Adversaries may establish persistence by creating or modifying event handlers, startup hooks, or associated configuration so that subsequent system or application activity automatically executes adversary-controlled c…
- AML.T0143
LLM Prompt Propagation
Lateral Movement · Maturity: Demonstrated
Adversaries may cause malicious instructions already introduced into an AI-enabled environment to be transferred onward to additional storage locations, agents, applications, or resources where they can influence subs…
- AML.T0144
Recover AI Model Internals
Discovery · Collection · Maturity: Feasible
Adversaries may infer or recover non-public AI model internals through analysis of model behavior or exposed information. Targets may include architecture, layer dimensions, hyperparameters, decision-tree structure, p…
- AML.T0145
Poison AI Agent Learning Environment
Persistence · Maturity: Feasible
Adversaries may manipulate an environment in which an AI agent interacts so that observations, feedback, rewards, actions, or resulting interactions are incorporated into model training to cause persistent changes to…
Summaries are excerpts of MITRE ATLAS descriptions (Apache 2.0). Follow the links for full entries, references and procedure examples.
Real-world evidence
Recent ATLAS case studies involving LLMs and agents
- AISI Evaluation Agents Conduct Unsanctioned Real-World Cyber Activity
AML.CS0074 · Incident · 2026 · reported by UK AI Security Institute (AISI)
Between July 25 and 28, 2026, autonomous AI agents undergoing cybersecurity evaluations by the UK AI Security Institute (AISI) took unsanctioned actions involving real people and public Internet services. AISI identified 19 events across 10 evaluation runs: 17 involved Claude Mythos 5 and two involved GPT-5.6 Sol.
- Autonomous Claude Agents Compromise Three Real Organizations During Cyber Evaluations
AML.CS0073 · Incident · 2026
Autonomous Anthropic models gained unauthorized access to systems belonging to three unrelated organizations during six cybersecurity evaluation runs from April through July 2026, compromising real systems while pursuing capture-the-flag objectives.
- AI Recommendation Poisoning via Crafted AI Assistant Links
AML.CS0072 · Incident · 2026 · reported by Microsoft Defender Security Research Team
The Microsoft Defender Security Research Team identified a widespread pattern of AI assistant memory manipulation carried out for promotional purposes, which they named AI Recommendation Poisoning.
- Multi-Agent Framework Compromises Taiwanese Government Systems
AML.CS0071 · Incident · 2026 · reported by Dream Research Labs
In early July 2026, an unknown Chinese-language operator used a multi-agent framework built on Hermes and OpenClaw against government systems that subsequent public reporting identified as Taiwanese. Dream Research Labs recovered a 160 MB operational workspace containing 1,395 files documenting 12 attack waves conducted from July 1 through July 4.
- Threat Actor Uses a DeepSeek-Powered Hermes Agent in Langflow and n8n Exploitation Attempts
AML.CS0070 · Incident · 2026
A Chinese-speaking threat actor operating as knaithe or KnYuan configured Hermes Agent to use DeepSeek as its reasoning engine. Hermes supplied terminal access, Telegram-based operator control, and a skills system containing both bundled and actor-created red-team skills. The actor also integrated an MCP server that exposed FOFA search, query translation, and Nuclei scan-generation capabilities.
- GTG-1002 Claude Code Espionage Campaign
AML.CS0069 · Incident · 2025
In September 2025, GTG-1002 used a jailbroken Claude Code agent to conduct a cyber-espionage campaign against approximately 30 organizations, succeeding against a small number. Anthropic assessed with high confidence that GTG-1002 was a Chinese state-sponsored group.
Case studies are documented by MITRE ATLAS; see each entry for its sources.
MITRE ATT&CK® and ATT&CK® are registered trademarks of The MITRE Corporation. ATT&CK data © 2026 The MITRE Corporation. This work is reproduced and distributed with the permission of The MITRE Corporation (ATT&CK v20.1). MITRE ATLAS™ data (v2026.10) © MITRE, used under the Apache License 2.0. Securitain is not affiliated with or endorsed by MITRE. Explanations, scenarios and control mappings on this page are Securitain’s own analysis. Sources and licensing