Teachers warn Leaving Cert changes risk inequality and AI misuse

Teachers warn Leaving Cert changes risk inequality and AI misuse, new ASTI survey finds. File picture

Irish teachers are “deeply dissatisfied” with changes to how the Leaving Cert is graded and say they will have extreme difficulty confirming whether AI has been used by students to achieve a higher mark.

New research from teaching union ASTI shows that the vast majority of teachers covering the four subjects currently utilising an Additional Assessment Component (AAC) believe that the 40% weighting given to project work is at least double what it should be.

ASTI president Padraig Curley said the proliferation of AACs across Leaving Cert subjects, as part of the redesign of the senior cycle curriculum, will “bake in” inequality into the system, as wealthier households are better positioned to manage the volume of project work involved.

A new survey by Red C for the ASTI shows that 82% and 83% of teachers of biology and chemistry describe themselves as either dissatisfied or very dissatisfied with allocating 40% of marks in those subjects to project work from next summer’s Leaving Cert.

Most teachers of the relevant subjects believe that 20% or less would be a “more appropriate” level for allocating final marks to non-exam coursework.

AACs were introduced — initially this year in four subjects: biology, chemistry, business, and physics, with more to follow in 2027 — as a means of easing the burden on students of having a year’s work assessed by a single exam.

However, concerns about how these AACs work in practice include the standard of infrastructure in schools to support practical project work, particularly in terms of safety, and the added stress for students of completing multiple highly weighted projects.

Separately, John Conneely, assistant general secretary with the ASTI, described the proliferation of AI — and teachers being unable to verify its use — as a “serious problem”.

“If there is no reliable way to verify authorship then there’s a real concern about the integrity of the assessment itself,” he said.

Mr Conneely said the issue is “a very, very difficult one”, noting that a student who utilises an AI model to compile work and then slightly edits the results would make the origin of that work practically undetectable.

These issues will be debated at the ASTI’s annual conference next week, with general secretary Kieran Christie noting that pay for teachers will also be a key issue, while refusing to rule out emergency motions on remuneration levels given ongoing global turmoil and the cost-of-living crisis.

“We certainly wouldn’t restrain ourselves,” Mr Christie said of the possibility the current motion for a 6% pay rise for teachers could prove insufficient given the levels of inflation currently being seen.

Negotiations for a new public pay deal are set to take place between unions and the Government later this year.