Assessing the causes of underperformance in artificial intelligence in education policy

The Beijing Consensus adopted in May 2019 provides the immediate context for the causes of underperformance in artificial intelligence in education policy. Corrective action concerning assessing the intended improvement should address the identified cause, assign responsibility and set a review period. Residual risk should remain open until sustained improvement is demonstrated.

For artificial intelligence in education policy, the Beijing Consensus on Artificial Intelligence and Education was adopted in May 2019. It addresses policy planning, management, teaching, learning, skills, lifelong learning, inclusion, gender equality, data and research. It promotes human-centred and equitable use rather than technology adoption as an end in itself. Authorities and providers should therefore connect each proposed use to an educational purpose, governance responsibility and evidence of benefit and risk.

Improvement objective and baseline

The formal status of the Beijing Consensus adopted in May 2019 should be preserved in any public account. For the causes of underperformance in artificial intelligence in education policy, the instrument should be used to identify the intended direction, the actors addressed and the implementation measures that remain necessary.

The position on assessing the corrective action should be established through proportionate evidence and should remain open to correction when material new information becomes available. Responsibility for corrective action should be identifiable at each consequential decision point. In work concerning artificial intelligence in education policy, delegation should identify both the operating role and the body retaining oversight of learner impact. The allocation of responsibility should prevent gaps between system oversight and institutional operation.

The review method for the corrective action should connect the question under examination to suitable evidence and a conclusion no broader than the tested scope. The evidential basis for the relevant practice should identify source, period, coverage and material limitations. For decisions concerning artificial intelligence in education policy, corroboration is required where a single record cannot support the decision. End-to-end assurance is required because individual functions may operate as designed while the combined process fails.

Governance of the intended improvement requires a clear allocation of authority, information and follow-through. For artificial intelligence in education policy, material matters should be referred to the body authorised to act or accept residual risk. Improvement work on corrective action should begin with a verified problem, defined baseline and measurable outcome. Completion should depend on evidence of effect rather than completion of planned activity. A decision should not be closed at the operating level where material impact, conflict or a significant evidential gap remains unresolved.

Assurance concerning the relevant practice should state the scope examined, evidence relied upon and any condition preventing a complete conclusion. Unsupported elements should remain open. In the present context, unequal performance across learner groups, unclear responsibility between providers and suppliers and unverified outputs entering teaching or assessment may produce acceptable aggregate reporting while individual learners remain exposed to material disadvantage. For decisions concerning artificial intelligence in education policy, adverse cases should form part of the sample wherever they may reveal a material control weakness.

The record for the relevant practice should identify the responsible function, decision authority and escalation route. In work concerning artificial intelligence in education policy, evidence outside the relevant period or scope should be identified and given no more weight than its limitations permit. An unresolved contradiction is a limitation on the conclusion and should be reported as such.

Controls and accountable action

The principal risks associated with assessing the matter should be assessed as connected conditions. Within the scope under review, adverse cases and unresolved contradictions should be retained because they may reveal limitations concealed by an average result.

Data used for the matter should be interpreted against stable definitions and an identifiable population. For artificial intelligence in education policy, changes in method, definition or series should remain separate from changes in the underlying result. Reporting should distinguish work performed from the outcome demonstrated after implementation. Closure reporting should not obscure unresolved action or risk retained by the responsible authority.

The assurance record for the intended improvement should permit another competent reviewer to understand the evidence, method, judgement and treatment of material exceptions. A conclusion on the corrective action should extend no further than the available evidence permits. For artificial intelligence in education policy, missing populations, inconsistent records and unresolved exceptions should be reported with the finding. Accuracy measured in one setting may not transfer to another population, language, curriculum or decision context. Any indicator used in relation to assessing the causes of underperformance in the corrective action should distinguish description from causal explanation. Interpretation should retain uncertainty, distributional differences and limits on generalisation. Data used for the intended improvement should be interpreted against stable definitions and an identifiable population. A revision or break in series should not be reported as a change in performance. A finding should not be separated from limitations capable of changing how it is understood or applied.

Governance of the matter requires a clear allocation of authority, information and follow-through. When examining artificial intelligence in education policy, traceable source and version information allow genuine improvement to be distinguished from administrative revision. The evidential history should preserve conclusions that were operative when a material decision was made.

The record for the corrective action should identify the responsible function, decision authority and escalation route. In work concerning artificial intelligence in education policy, relevant evidence should reach the body authorised to commit resources, amend policy or accept residual risk, and its judgement should be recorded.

For corrective action, the implementation record should distinguish binding duties, policy expectations and institutional choices, including any transition or jurisdictional limitation. Where evidence concerning artificial intelligence in education policy cannot support assurance, the limitation should be reported and corrective work should remain open.