Sydney schools ordered to retake NAPLAN after predictive text enabled during test
Year 5 students at two Sydney schools were required to retake their NAPLAN writing tests after Apple's predictive text feature was inadvertently enabled during the assessment. Schools reported the issue to regulators, and students completed the retake with a new prompt, though one school faced delays obtaining necessary access codes.
What happened
Year 5 students at Waverley College and Kambala in Sydney were required to retake their NAPLAN writing tests after Apple's predictive text and spellcheck features were inadvertently enabled during the assessment.
A teacher at Kambala and students at Waverley College discovered the issue when they noticed the feature was active. Waverley College immediately reported the problem to the NSW Education Standards Authority (NESA), and Kambala reported to both NESA and the Australian Curriculum and Reporting Authority (ACARA).
Students were given a new prompt for the revised test. Waverley College experienced delays accessing the rescheduled test due to missing student profile codes, and students ultimately completed the retest on Monday.
Reported impact
The predictive text feature, which uses artificial intelligence to suggest word completions and generate responses based on a user's writing style, potentially affected the validity of students' original test submissions.
Waverley College noted that other schools may have experienced the same technical difficulty without realizing predictive text was enabled during their students' assessments.
Index analysis
ACARA reported that more than 880,000 online writing tests had been submitted successfully in the first four days of testing in 2025, indicating the issue was limited to a small number of schools.
Schools had been provided with instructions to disable predictive text before testing and were reminded again on the second day of testing by NESA.
Governance controls
NESA stated that schools are provided with measures to maintain test integrity, including instructions to disable predictive text on devices before testing begins.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.