ACTis provides risk analysis and early intervention recommendations against state capture and corruption in South African government institutions.
The 9 ACTis Flags
Flag #1
Improper appointments
Flag #2
Co-opting key functions
Flag #3
Climate of fear
or favour
Flag #4
Corrupt
awards
Flag #5
Financial malpractice
Flag #6
Obstructing accountability
Flag #7
Manipulating perceptions
Flag #8
Private
gain
Flag #9
Shadow
network
Click to view reports by flags raised.
Latest Status Reports
All Reports
Status Reports
National Government
Provincial Government
Municipal Government
Public Entities
Constitutional Entities
More about ACTis
What is ACTis?
The Anti-Corruption Tracker Information System is an initiative of the Anti-Corruption and Accountability Project at the Institute for Security Studies. It provides analysis on corruption and state capture in South Africa, to inform early intervention and enforcement. It utilises a system of indicators and an AI-driven data-collection and analysis tool, validated by human experts. It identifies risks within individual government institutions and produces actionable recommendations to address them, with a view to disrupting patterns of state capture and corruption. Findings are shared directly with government and other duty-bearers, civil society organisations, the media and the public, empowering authorities to intervene, strengthening accountability, and enabling citizen activism.
Analytical Framework
The Analytical Framework sets out the thinking, methodology and evidence base that underpin ACTis. It draws primarily on the findings and analysis of South Africa’s Zondo Commission, informed by relevant international approaches to measuring corruption and state capture, while developing a methodology specifically suited to the South African context. ACTis aims to generate credible, institution-level evidence on risks of corruption and state capture, identify emerging patterns and networks, and translate findings into practical recommendations that enable those with the authority to act to intervene earlier. The methodology combines publicly available data from credible sources with a custom-built AI-assisted research tool and human analysis. Data is assessed against a structured set of individual and composite indicators, with findings synthesised into reports that identify risks, explain their significance and highlight potential points for intervention. ACTis does not seek to prove corruption or state capture conclusively; rather, it provides evidence-based risk indications to support investigation, accountability and timely action.
Flag Methodology
The Flag Methodology explains how ACTis converts publicly available information into structured indicators of corruption and state capture risk. It provides the detailed indicators, thresholds and rules used to move from published evidence to signs, indicators and ultimately flags. In brief, raw data from credible government, legal, parliamentary, civil society and media sources is processed through the ACTis AI-assisted system, which identifies and tags relevant “signs” in the underlying text. Human analysts then verify the sources and assess whether the evidence meets the criteria for individual indicators. These binary indicators are aggregated according to defined thresholds and rules to determine whether one of ACTis’ nine composite indicators, or “flags”, should be raised. A raised flag indicates that a particular pattern associated with corruption or state capture may be present and warrants further investigation or action.
Terms and Definitions
Further project information
About the Institute for Security Studies
Contact us: actis@issafrica.org
Disclaimer: This report is produced in the public interest and is based on official public records and other reputable publicly available sources considered reliable at the time of publication. The ISS has taken reasonable care to verify the information presented. Any analysis or opinions expressed are based on the available information and do not constitute findings of criminal, civil or administrative liability.