NSW moves to protect AI academic integrity in HSC assessments

NSW moves to protect AI academic integrity in HSC assessments

NSW schools have been directed to halt take-home assessments while the state government reviews the impact of artificial intelligence on student learning. The decision puts AI academic integrity at the centre of a national conversation for school leaders and system administrators.

Deputy Premier and Minister for Education and Early Learning Prue Car has asked the NSW Education Standards Authority (NESA) to urgently examine measures to protect the Higher School Certificate (HSC) as AI use accelerates in classrooms.

As an immediate step, NESA has been asked to consider a moratorium on unsupervised, take-home assessment tasks. A broader review of AI’s impact on student learning will follow.

For principals and curriculum leaders, the directive signals a shift in how governments view assessment design. It has moved from being a purely pedagogical choice to a governance responsibility.

AI academic integrity and the HSC under scrutiny

Half of a student’s HSC result is based on school-based assessment tasks. A portion of those tasks is currently completed outside the classroom. That unsupervised component is what has drawn the government’s attention, given how readily AI tools can generate assessment-ready content.

The review will examine:

  • AI’s effect on students’ cognitive development
  • how assessment methods can better safeguard authentic student output
  • whether major projects and creative works require specific protections

NESA has also been tasked with developing a common approach to identifying AI misuse. It will help teachers spot inappropriate AI use in submitted work.

Subject to NESA’s advice, interim changes would take effect by the start of Term 4, 2026. That gives the HSC Class of 2027 and their schools clarity ahead of their final year. Some tasks, such as major artworks and design and technology projects, are expected to be exempt.

“If students rely on AI to bypass the learning process, we risk undermining both their education and the integrity of the HSC,” Minister Car said.

What this means for school leaders

The NSW announcement arrives as Australian education providers acknowledge they are struggling to keep pace with AI’s spread. According to ReadyTech’s AI and Innovation Benchmark 2026, 43.8% of education organisations expect AI-related academic integrity issues to worsen over the next two years.

Despite those concerns, only 34.4% of institutions have a documented AI use policy. That governance gap will likely draw greater scrutiny following the NSW move.

School leaders will need to act on two fronts:

  • assessment redesign to reduce unsupervised, take-home components
  • staff capability to detect and respond to AI misuse

NESA’s forthcoming detection guidance will be one input. But principals who have already worked through building a structured AI safety approach for their school will be better placed to act quickly when formal advice arrives.

The cognitive aspect of this debate also deserves attention. Researchers have argued that outsourcing thinking to AI impairs the development of skills that assessment is designed to measure.

A June 2026 analysis examined what human cognitive abilities AI still cannot replicate. That research is directly relevant to how school leaders frame assessment’s purpose in staff and parent conversations.

“Technology should enhance education, not diminish it. Parents deserve confidence that HSC results reflect a student’s own knowledge, effort and achievement,” Car said.

Working towards AI academic integrity: What comes next

School leaders should anticipate NESA guidance before Term 4. There are practical steps to take now:

  • Review existing assessment calendars for unsupervised take-home tasks
  • Audit which subjects may qualify for an exemption
  • Brief staff on the likely direction of NESA’s detection framework

The broader question for education executives is whether NSW’s intervention becomes a national benchmark. Other states are watching. If NESA’s framework produces workable detection guidance, pressure will build on curriculum authorities elsewhere to follow.

AI academic integrity is no longer a future concern to be managed incrementally. For principals across Australia, it is an immediate operational and governance challenge that demands a position and a plan.