Artificial Intelligence (AI) System
A machine-based system that makes predictions, recommendations, or decisions influencing environments based on human-defined objectives.
How the laws define this
Regulators consistently treat these systems as technology that processes data or analyzes an environment to generate outputs, make decisions, or perform tasks simulating human intelligence such as learning and problem-solving to influence real, physical, or virtual environments. Key divergences center on how broadly instruments define a system's baseline nature and goals. While Kigali Global AI Summit Outcomes (Africa Declaration on Artificial Intelligence) and National Security Presidential Memorandum/NSPM-11 — Artificial Intelligence in the National Security Enterprise explicitly restrict definitions to machine-based systems operating for human-defined objectives, 인공지능(AI) 보안 안내서 broadens goals to include explicit or implicit objectives. Crucially, Cybersecurity Framework Profile for Artificial Intelligence (NISTIR 8596) departs from pure machine requirements altogether by encompassing both engineered and natural systems that learn, reason, or adapt to generate outputs for given inputs.
Synthesised from the 8 statutory definitions below. Each one is quoted in full, with its source.
How jurisdictions differ
What is distinctive in each definition, binding law in force first.
| Jurisdiction | Instrument | Status | What is distinctive |
|---|---|---|---|
| Colorado, United States | Concerning restrictions on the use of artificial intelligence relating to psychotherapy services. | In Force | Limits AI to tasks typically requiring human intelligence like learning, problem-solving, and decision-making, specifically in psychotherapy services context. |
| United States | National Security Presidential Memorandum/NSPM-11 — Artificial Intelligence in the National Security Enterprise | In Force | Adds requirements for AI systems to be reliable, robust, steerable, controllable, and compliant with laws and policies in national security enterprise. |
| United Nations | Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development | In Force | Emphasizes systems displaying intelligent behaviour through environmental analysis and autonomous actions to achieve goals, including knowledge-based, machine learning-based, or hybrid types. |
| South Korea | Artificial Intelligence (AI) Security Guide | In Force | Specifies that AI infers how to generate outputs for explicit or implicit objectives, influencing physical or virtual environments. |
| China | Ethics-Safety Guidelines for Artificial Intelligence Applications 1.0 | In Force | Widens to intelligent systems with autonomous perception, memory, decision-making, interaction, execution, and integration with cyberspace and physical world via advanced technologies. |
| Vermont, United States | Vermont DFR Bulletin No. 229 — The Use of Artificial Intelligence Systems in Insurance | In Force | Focuses on advanced technologies simulating human intelligence for insurance functions like underwriting, pricing, claims, and fraud detection affecting consumers. |
| Rwanda | Kigali Global AI Summit Outcomes (Africa Declaration on Artificial Intelligence) | Adopted | Adds that AI systems operate with varying degrees of autonomy and adapt behavior by interpreting complex data. |
| United States | Cybersecurity Framework Profile for Artificial Intelligence (NISTIR 8596) | Draft | Broadens to include both engineered and natural systems that generate outputs from inputs with abilities to learn, reason, or adapt. |
What it triggers
- In Colorado, mental health providers are prohibited from using artificial intelligence for direct therapeutic communication or unsupervised treatment.source
- In Colorado, mental health providers must ensure human oversight and informed consent when using artificial intelligence.source
- Under NSPM-11, AI systems in the national security enterprise are expected to be reliable, robust, steerable, and controllable.source
Definitions (8)
An Artificial Intelligence (AI) System refers to a machine-based system that, for a given set of human-defined objectives, can make predictions, recommendations, or decisions influencing real or virtual environments. These systems are designed to operate with varying degrees of autonomy and can adapt their behavior by interpreting complex data.
A machine-based system that, for explicit or implicit objectives, infer how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.
A system designed to perform tasks that typically require human intelligence, such as learning, problem-solving, and decision-making, as referenced in the context of psychotherapy services.
Within the national security enterprise, an AI system refers to any machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments. These systems are expected to be reliable, robust, steerable, and controllable, operating in accordance with applicable laws and policies.
Refers to intelligent systems capable of autonomous perception, memory, decision-making, interaction, and execution, rapidly integrating with cyberspace and the physical world, especially with the emergence of advanced technologies like large language models.
Advanced analytical and computational technologies that enable machines to simulate human intelligence, including learning, problem-solving, and decision-making capabilities. In the insurance context, these systems are applied across various functions such as underwriting, pricing, claims processing, and fraud detection, influencing or making decisions affecting consumers.
Refers to systems that display intelligent behaviour by analysing their environment and taking actions – with some degree of autonomy – to achieve specific goals. These systems can be knowledge-based, machine learning-based, or hybrid.
Broadly refers to engineered or natural systems that generate a particular output for a given set of inputs, with the ability to learn, reason, or adapt. This is a foundational concept for the entire framework.
Not to be confused with
Related term 'AI System' simply describes a machine-based system inferring outputs from inputs, lacking the detailed human objectives and environmental influence in most definitions here.
Related term 'Artificial Intelligence System (AI System)' mirrors core machine-based inference for predictions or decisions but omits specifics like autonomy or context found here.
Questions and answers
What restrictions apply to AI in mental health under Colorado law?
Colorado HB 1195 prohibits mental health providers from using AI for direct therapeutic communication or unsupervised treatment. It mandates human oversight and informed consent. The Act took effect on August 12, 2026.
What qualities must AI have in US national security?
NSPM-11 requires AI systems to be reliable, robust, steerable, controllable, and operating in accordance with applicable laws and policies within the national security enterprise.
Does AI include natural systems in cybersecurity frameworks?
NISTIR 8596 broadly defines AI to include both engineered and natural systems capable of learning, reasoning, or adapting for given inputs.
Related Terms
AI System
Machine-based system that infers outputs from inputs....
Artificial Intelligence Systems
Systems capable of perceiving an environment, interpreting information, taking action, and learning by analyzing prior actions....
AI (Artificial Intelligence System)
Any software, machine, or system simulating humanlike cognitive functions like learning or reasoning....
System based on AI
An electronic system that makes predictions, recommendations or decisions....
Artificial Intelligence System (AI System)
A machine-based system inferring from input to generate outputs like predictions or decisions influencing environments....