evidoria

← Back to browse

Good practice Imported

Curtin University Disables Turnitin's AI-Writing Detection — An Equity-Driven Policy Reversal

Australia · Perth · See the Australia profile · See the Perth profile

Evidence: Descriptive / self-reported Top 40% 53/100 · Ask Evidence Copilot about this practice

From 1 January 2026, Curtin University (Australia) disabled Turnitin's AI-writing detection campus-wide, keeping only text-matching checks, citing reliability problems and unequal false-positive rates across student groups — joining dozens of universities abandoning AI detectors.

Curtin University Disables Turnitin's AI-Writing Detection — An Equity-Driven Policy Reversal

Details

Promoter
Curtin University
Period
Announced 4 September 2025; effective 1 January 2026
Keywords
higher education, academic integrity, edtech policy, generative AI

Context

Turnitin's AI-writing detector has been used by universities worldwide to flag student work suspected of being generated by tools like ChatGPT, but the feature has drawn sustained criticism from academics over its accuracy, including calls for public access to the detection algorithm for independent verification.

Activities

On 4 September 2025, Curtin University announced that from 1 January 2026 it would disable Turnitin's AI-writing detection feature across all campuses and study periods, while keeping Turnitin's separate text-matching originality checks in place, citing equity concerns about detection tools flagging some student populations more than others.

Results

Curtin's announcement provided no statistics of its own on false-positive rates or affected student numbers, a gap it does not hide; Turnitin itself has publicly acknowledged that no AI system can achieve zero false positives. Curtin is one of more than 50 universities reported to have scaled back or disabled Turnitin's AI-detection feature, alongside institutions such as Johns Hopkins, Vanderbilt and Waterloo.

Conclusions

The case is a useful, honestly-documented example of an institution choosing to walk back an AI tool rather than expand one, framing the decision as fostering trust and ensuring assessments remain fair and future-ready.

Implementation
Indicative cost
Low (< €50k)
Time to results
Short (< 1 year) — Announced 4 September 2025; effective from 1 January 2026, roughly a four-month lead time.
Staffing & skills
university administration (policy decision, no dedicated delivery staff described)

Conditions for success

  • clear public communication of rationale and effective date to the university community
  • retaining text-matching originality checks while disabling only the AI-detection feature, avoiding an abrupt full withdrawal from Turnitin

Common failure modes

  • no disaggregated false-positive statistics were published to substantiate the equity rationale

Where it fits

Governance type
university administration
Scale
single institution, all campuses and study periods
Income level
high-income

Commonly funded by

Own resources / municipal budget National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful