Ceibal & ANEP's AI-Assisted Basic-Cycle Accreditation Exam
Uruguay
Uruguay's Ceibal and ANEP built an AI grader for the basic-cycle accreditation exam taken by roughly 8,000 mostly out-of-school young …
United Kingdom · Cambridge · See the United Kingdom profile · See the Cambridge profile
Evidence: Descriptive / self-reported Top 3% 80/100 · Ask Evidence Copilot about this practice
Free AI writing-feedback tool by Cambridge University Press & Assessment (Sept 2016); assigns CEFR A1–C2 levels to EFL essays and highlights errors by type; grounded in Yannakoudakis et al. (2018) peer-reviewed model and the open Write & Improve Corpus 2024 (23,000+ essays).
Write & Improve is a free automated writing-feedback tool developed by Cambridge University Press & Assessment, launched publicly in September 2016 after more than three years of beta testing. It assigns CEFR proficiency levels (A1-C2) to submitted English texts and highlights errors by type (grammar, vocabulary, spelling), using a model trained on the 30-million-word Cambridge Learner Corpus and validated in peer-reviewed research (Yannakoudakis et al., 2018). The tool is free with no mandatory account for basic use and is used for self-study and in EFL classrooms across 100+ countries.
A 2025 study published in Language Teaching (Cambridge Core) found AI-to-human-rater agreement on CEFR level comparable to inter-rater reliability between two trained human examiners. In 2024 Cambridge released the Write & Improve Corpus 2024, an openly licensed dataset of 23,000+ annotated essays (3.5 million+ word tokens) from 766 learners, enabling further research benchmarking.
Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.
Own resources / municipal budget
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.
Where this practice's information was retrieved from, and when.
Uruguay
Uruguay's Ceibal and ANEP built an AI grader for the basic-cycle accreditation exam taken by roughly 8,000 mostly out-of-school young …
Kazakhstan
A quasi-experimental study at a private Kazakhstani university gave 134 trainee English teachers either AI feedback (Cambridge's Write & Improve) …
China
A 12-week RCT with 190 Chinese university EFL students found Grammarly-based automated writing evaluation significantly improved task achievement, coherence, lexicon …
Australia
Instead of an unwinnable AI-detection arms race, Sydney redesigned assessment into two lanes — secure in-person assessment of core capability, …
Open full copilot Grounded in cited practices — always check the sources.