Why not all SIU referral results in prosecutions

The National Prosecuting Authority (NPA) has sought to explain why 5,944 criminal referrals made by the Special Investigations Unit (SIU) since 2011 have translated into a much smaller number of criminal dockets and prosecutions.

The special director of public prosecutions who heads the NPA Specialised Commercial Crime Unit, Nkebe Kanyane, told parliament’s standing committee on public accounts (Scopa) yesterday that the figures could not be compared on a one-to-one basis because multiple referrals can be consolidated into a single criminal investigation.

The SIU has made 5,944 referrals to the NPA since 2011, with the number increasing sharply in recent years. But only 10% translated to dockets.

There were 901 referrals in 2024 and 899 in 2025, compared with 20 in 2011 and 23 in 2012.

Kanyane said the process begins when the president issues a proclamation authorising an SIU investigation. Where the SIU uncovers evidence pointing to a criminal offence, it refers the matter to the NPA, which assesses the referral and can return it to the SIU or refer it for further investigation.

Depending on the nature and complexity of the case, investigations can be conducted by the Hawks, the SAPS or other specialised investigative environments. The NPA can also investigate certain matters itself.

Kanyane said several referrals could relate to the same individuals or conduct and be dealt with in a single docket. “The statistics/numbers will never tally,” she said.

Gauteng has accounted for 2,655 referrals since 2011. Those resulted in 206 registered dockets, of which 119 were under Hawks investigations, 40 were at the decision or guidance stage and 12 were in court.

KwaZulu-Natal had 1,045 referrals, resulting in 29 registered dockets, with 10 under investigation, 11 at the decision or guidance stage and four in court.

In the Eastern Cape, 871 referrals resulted in 84 registered dockets, of which 13 were under investigation, 24 were at the decision or guidance stage and 30 were in court.

Kanyane said the gap between referrals and prosecutions also needed to be understood against the complexity of corruption cases. Investigations often began years after the alleged conduct, making it difficult to locate witnesses and obtain documentary and electronic evidence.

Electronic evidence had to be downloaded and analysed, while evidence located outside SA often required formal mutual legal assistance processes. Securing forensic auditors could also take considerable time, while the state continued to lose experienced investigators and prosecutors to the private sector.

The NPA also lacks an integrated justice-sector case-management system to track matters from investigation and referral through to prosecution and asset recovery.

Cases can be declined where there is no prima facie evidence of criminality, insufficient prospects of a successful prosecution or insufficient evidence to prove the relevant offence. Difficulties securing witnesses or documents can also affect prosecutorial decisions.

The SIU said contracts worth R38.1bn had been set aside as a result of its investigations, while R2.9bn had been recovered by the SIU and a further R8bn by the Asset Forteiture Unit (AFU), taking combined recoveries to R11.5bn.

In an Eskom-SAP matter, two contracts worth R1.2bn were set aside and SAP was ordered to repay R570m, which was subsequently repaid to Eskom. The SIU made 13 criminal referrals to the NPA.

In a separate Eskom matter involving ABB, the NPA said it had reached an agreement with the company for more than R2.5bn in punitive reparations in connection with corruption-related activities during the state-capture period. The matter remained under investigation and had also been referred to the AFU.