Fraud and corruption control evaluating compliance and its drivers (November 2023)
- Report length
- 32 pages
Summary
The report's own summary: Chapter 2: Overview of compliance with the circular, pages 8 to 10
Generated summary: written by an AI model from the report text and checked against the cited pages.
The NSW Independent Commission Against Corruption assessed agencies’ compliance with the Treasury fraud and corruption control policy and barriers to meeting it.
Its research combined survey responses from 90 agencies, nine focus groups and analysis of agencies’ control documents.
The Commission reports that most agencies met minimum standards, while managing third parties was the requirement agencies found hardest to meet.
The Commission advises agencies to improve third-party controls through due diligence, clear responsibilities, staff training and reporting arrangements.
In the body's own words
- Finding
The research findings show that many agencies do not consistently apply third-party controls.
Findings and recommendations
Allegations are not findings. A commission's finding or opinion of corrupt conduct or misconduct is not a finding that a person is guilty of a criminal offence, and the commissions say so in their reports.
No findings or recommendations are reproduced in this record. This does not establish that the report contains none.
Published responses
Responses to this report have not been checked.
Notes on this record
- Corruption prevention and research reports make no findings about individuals; any recommendations are not extracted unless set out as numbered recommendations that could be read confidently.
Coverage
This publisher
- Reports on investigations made public before 2016 are not online: the ICAC keeps only a summary for each, and copies must be requested from the ICAC. Those records hold the ICAC's published summary; findings and recommendations are not extracted.
- The ICAC copyright statement reserves all rights (fair dealing aside). OPAX keeps short verbatim finding and recommendation passages with links to the original report.
- Findings are read from the summary chapter of each report PDF. Where a finding's particulars run past the cited page only the finding statement is kept, and reports whose layout could not be read confidently are marked not_extracted.
- Agency action plans and progress reports are listed as documents; their contents (implementation status for each recommendation) are not parsed.
- Report PDFs without a text layer (scanned images, for example Operation Scania) are not OCRed; their findings stay not_extracted.
- Many pre-2016 summaries give only the month a report was made public; those records carry no date and a note with the month, because the schema holds day-level dates only.