Technical

AI solution

System using ML, statistical inference, or automated decision tech.

How the laws define this

Regulators consistently recognize an AI solution as a system, software, or service that utilizes automated techniques, such as machine learning, to execute tasks, process data, or support decision-making. However, these instruments diverge significantly in how they delineate their scope. The FAIRA guideline (Queensland Government) adopts a broad, perception-based approach, covering any function stakeholders reasonably regard as AI to prevent rigid thresholds. In contrast, other frameworks define scope through technical methods or specific domains. For instance, Portaria CNJ nº 270/2025 explicitly restricts its definition to technologies processing data for judicial activity, while the Seattle Responsible AI Policy confines its application to tools acquired or operated to provide municipal services or address City challenges. Meanwhile, instrument frameworks like the AI Action Plan (Kratt) 2022–2023 and Developing Artificial Intelligence Solutions (National AI Strategy) maintain a broader purview across public and private sector tasks without restricting the term to a single specialized domain.

Synthesised from the 5 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. Showing 4 of the 5 instruments that define this term; all are listed under Definitions below.

Jurisdiction	Instrument	Status	What is distinctive
Brazil	Portaria CNJ No. 270/2025	In Force	Limits AI solutions to those used specifically in judicial activity, requiring registration, classification, and governance under the CNJ framework.
Queensland, Australia	Foundational Artificial Intelligence Risk Assessment (FAIRA) guideline (Queensland Government)	In Force	Broadens scope to any product, service, or system perceived as AI by stakeholders, avoiding rigid thresholds for risk assessments.
Washington, United States	Seattle Responsible AI Policy	In Force	Confines AI solutions to those acquired, configured, developed, operated, or maintained explicitly for municipal services or city challenges.
Latvia	Developing Artificial Intelligence Solutions	Adopted	Encompasses deployed AI applications and services across both public and private sectors using techniques like machine learning and NLP.

What it triggers

Under Portaria CNJ No. 270/2025 in Brazil, AI solutions used in judicial activity must be registered, classified, and governed.source
The Queensland FAIRA guideline recommends AI solutions undergo risk assessment across the AI lifecycle for public sector agencies.source
Seattle's Responsible AI Policy guides the ethical use of AI solutions in municipal services, focusing on equity, transparency, and risk mitigation.source
Latvia's national AI strategy sets out objectives for developing AI solutions across public and private sectors.source

Definitions (5)

Any system developed or deployed using machine learning, statistical inference, or other automated decision‑support technologies to process data and produce outputs used in judicial activity. The term frames which tools must be registered, classified and governed under the CNJ framework.

Portaria CNJ nº 270/2025 (Portaria designating the members of the National Committee on Artificial Intelligence of the Judiciary - Comitê Nacional de Inteligência Artificial do Judiciário)Definition 1 of 5

Any software or service that employs automated learning, models or related techniques to provide decision‑support or automation in a given task; used to scope which projects, deployments and components fall within the Kratt activities.

AI Action Plan (Kratt) 2022–2023 (Tehisintellekti tegevuskava 2022–2023)Definition 2 of 5

An AI solution is any product, service or system that performs functions commonly described as artificial intelligence or that stakeholders would reasonably regard as AI. The FAIRA guideline uses this working definition to determine scope for assessment and to avoid rigid thresholds that might exclude systems perceived as AI by users or stakeholders.

Foundational Artificial Intelligence Risk Assessment (FAIRA) guideline (Queensland Government)Definition 3 of 5

Refers to systems utilizing Artificial Intelligence that are acquired, configured, developed, operated, and/or maintained to provide municipal services or solve City challenges.

Seattle Responsible AI PolicyDefinition 4 of 5

Refers to deployed artificial intelligence applications and services across public and private sectors that use methods like machine learning, pattern recognition, natural language processing and related automation techniques to support decision-making or automate tasks.

Developing Artificial Intelligence Solutions (National AI Strategy)Definition 5 of 5

Not to be confused with

AI system or service

AI system or service emphasizes algorithmic or statistical processes for decision support, while AI solution here often highlights broader applications or services in specific contexts like judicial or municipal use.

High-risk automated decision system

High-risk automated decision system focuses on systems with significant legal impacts in sensitive domains, whereas AI solution applies more generally without specifying risk levels.

Questions and answers

What triggers compliance for AI solutions in Brazilian courts?

AI solutions processing data for judicial outputs must be registered, classified, and governed under the CNJ framework via the National Committee on AI.

Does the Queensland guideline cover user-perceived AI?

Yes, it includes any product, service, or system reasonably regarded as AI by stakeholders to determine scope for risk assessments.

What AI solutions does Seattle's policy address?

Those acquired, configured, developed, operated, or maintained to provide municipal services or solve city challenges, with a focus on ethical guidelines.