NIR ACADEMY · RESEARCH & DEVELOPMENT
NIR AI-Native Language Competence Framework™
Rethinking Language Learning, Human Agency, and Responsible AI Use.
A proposed multidimensional framework exploring how language competence and AI-supported learning capabilities can be described, investigated, and evaluated in an AI-rich educational environment.
Beyond AI-Enhanced Language Learning
Many AI tools make familiar language learning faster or more convenient. NIR Academy’s AI-native language competence framework asks a further question: what should language competence mean when AI is part of everyday learning?
AI-enhanced learning keeps the goals of learning unchanged and uses AI to support them, for example by giving feedback or generating practice material.
AI-native learning starts from the assumption that AI is present. It treats the learner’s ability to use, question, and direct AI as part of what is learned and assessed, alongside language performance.
“AI-native” is NIR Academy’s project-specific working concept. It is not an established, universal definition.
Four Proposed Constructs of the AI-Native Language Competence Framework
The AI-native language competence framework proposes four constructs. All four are provisional and remain under investigation.
Language Competence
Language performance demonstrated under clearly specified task and assistance conditions.
Provisional · under investigationAI Literacy
Understanding, evaluating, and purposefully using AI tools.
Provisional · under investigationHuman Agency
Goal-setting, decision-making, self-direction, and ownership of learning.
Provisional · under investigationCritical and Responsible AI Use
Evaluating AI outputs, recognising risks and limitations, and making responsible decisions.
Provisional · under investigationHow the Constructs May Relate
Select a construct to highlight its description above. The dashed outline shows hypothesised links only.
- AIL may shape how effectively learners use AI to support LC.
- HA may influence how learners set goals and choose when to use AI.
- CRAI may draw on both AIL and HA when outputs are judged.
- LC may be demonstrated differently depending on the assistance allowed.
These relationships are proposals for investigation. Whether the four constructs are empirically distinct has not been established.
From Conceptual Design to Empirical Investigation
The research pathway for the AI-native language competence framework has six stages. The conceptual framework, preliminary construct specifications, and draft expert-review materials are under development. Expert review, pilot testing, and empirical validation are planned subsequent stages.
- Conceptual framework developmentUnder development
- Construct and descriptor specificationUnder development
- Expert reviewPlanned
- Descriptor and task refinementPlanned
- Pilot studyPlanned
- Further validation researchPlanned
No expert review, pilot study, or validation has been completed.
Relationship to Existing Frameworks
The proposed framework is being developed in dialogue with relevant research and established resources.
CEFR Companion Volume
Council of Europe descriptors for language proficiency, a reference point for describing language competence.
Visit the Council of Europe page (opens in a new tab)UNESCO AI Competency Frameworks
UNESCO frameworks describing AI competencies for students and teachers, relevant to AI literacy and responsible use.
Students framework (opens in a new tab) · Teachers framework (opens in a new tab)Testing Standards
The standards for Educational and Psychological Testing, which guide validity evidence, fairness, and test development.
Visit the official AERA Standards page (opens in a new tab)These organisations have not endorsed, reviewed, or validated the NIR framework. They are listed as sources the project engages with.
Research Documents and Resources
None of these documents is publicly available yet. Links will be added when each one is released.
Concept Note
Rationale and working definitions for the framework.
In developmentConstruct Specification and Descriptor Development
How the four constructs and their descriptors are being specified.
In developmentExpert Review Protocol and Evaluation Toolkit
Materials for structured expert review of constructs and descriptors.
In developmentFuture Research Publications
Papers and reports arising from the research pathway.
PlannedCurrent Status: Under Development — Pre-Validation
- The proposed constructs and descriptors have not yet been empirically validated.
- The N1–N5 progression remains a research hypothesis. It must not be read as established learner levels or as equivalents of CEFR levels.
- Content on this page may change as the research develops.
Advancing Research in AI-Native Language Learning
NIR Academy welcomes research dialogue and appropriate academic collaboration in language assessment, AI literacy, learner agency, and responsible AI in language education.
Explore NIR Research
