Dating App Kidnappings in South Africa: What the Evidence Shows
Dating App Kidnappings: What the Evidence Shows
Kidnappings in South Africa have risen 264% in a decade, and half of them happen in a single province. A small number begin somewhere the statistics cannot see: on a dating app. In July a court jailed four men for exactly that. It also heard that most of their victims had refused to lay charges at all.
1. Is Kidnapping Really Rising?
Yes, though not everywhere, not evenly, and mostly not in Cape Town.
Across South Africa, reported kidnappings went from 4,692 in 2014/15 to 17,061 in 2023/24. That is a rise of 264% in ten years. Parliament's own analysts found the figure had almost doubled since 2021/22 alone. Whatever else is disputed about South African crime numbers, this trend is not.
The question is where.
Where South Africa's Kidnappings Happen
Share of all reported kidnappings by province. Gauteng's share has been recorded between 51% and 54.8% depending on the quarter; the figure shown is the lower of the two. Source: SAPS Crime Registrar data, as analysed by the Institute for Security Studies.
So the plain answer to the question a Cape Town reader is likely to be asking: no, this city is not the centre of it. Gauteng records more than half of all kidnappings in the country. KwaZulu-Natal takes another fifth. The Western Cape sits at around 6%, level with Mpumalanga.
Two qualifications keep that from being the end of the matter.
The first is that the Western Cape's small share is a different kind of small. When the province last drew attention in these figures it came second in the country for kidnappings where a ransom was demanded, and third where the motive was extortion. Elsewhere, kidnapping is mostly something that happens during a hijacking. In the Western Cape, robbery accounts for 47% of cases, against nearly 80% in Gauteng. The remainder are something else.
The second is that Cape Town's own crime problem is severe and lies elsewhere. Four of the country's five worst police stations for murder are in the City of Cape Town district: Mfuleni, Nyanga, Delft and Gugulethu. Of 276 gang-related murders sampled nationally in one recent quarter, 257 were in the Western Cape. A Cape Town resident asking what is most likely to harm them is not asking about kidnapping.
How this article was reported, and what it does not claim
- Everything here is sourced
- Every figure, name, date and charge in this article comes from court reporting, SAPS quarterly statistics, parliamentary and legislature records, or the Institute for Security Studies. Sources are listed in full at the end. Where a claim is contested or unverifiable, it is labelled as such rather than smoothed over.
- Accused and convicted are different words
- Two of the three cases examined here are still before the courts. The men named in the Johannesburg and Cape Town matters are accused, not convicted, and are entitled to the presumption of innocence. Only the four men sentenced in Durban in July 2026 are described here as having committed crimes.
- The scale is genuinely unknown
- South Africa's crime statistics contain no category for dating-app-initiated kidnapping, and no category for crimes at short-term rental properties. Any number purporting to measure the phenomenon nationally, including estimates by private security firms, is an extrapolation. This article says so wherever it uses one.
- No operational detail
- This article describes the pattern of these crimes at the level established in open court and published warnings. It deliberately withholds nothing that would help a victim, and adds nothing that would help an offender.
2. The Case That Set the Pattern
He was 18, in his first year at the University of the Witwatersrand. In September 2023 he opened Grindr, agreed to meet someone and travelled to an address in Johannesburg. Nothing about the evening was remarkable. That is close to the whole point.
His roommate reported him missing. Within a day photographs of him reached his family, along with a demand for R30,000.
A family friend sent R3,000. It was the smallest sum in the case and the one that decided it. Detectives from Hillbrow police station took up position at the BP and Shell garages near Denver Hostel, a migrant labour hostel east of the city centre, and waited to see who would walk out to collect it. A man left the hostel and went to an FNB cash machine. He was carrying two phones. On one of them were the photographs that had been sent to the family.
The student was found alive. Seven men were arrested. In July 2026, thirty-four months later, their trial is still hearing prosecution witnesses. It resumes in August.
One detail from the investigating officer's testimony has stayed with people who follow these cases. He told the court that he had first come across the app being used in violent crime in May 2023. That was four months before the student was taken, and it was another matter entirely. The pattern was already running.
By the time the case reached a courtroom, versions of it had been recorded in KwaZulu-Natal and the Western Cape.
3. What These Numbers Actually Contain
The figure everyone quotes is 27 a day. It is correct, and it is usually quoted wrongly.
It comes from Gauteng: 2,452 kidnappings between January and March 2026, divided by ninety days. Not South Africa. Nationally the daily average has run far higher, around 50 in the July to September quarter of 2024.
And nothing spiked. Gauteng recorded 2,414 kidnappings in the same quarter a year before and 2,452 in this one. Thirty-eight more cases, a rise of 1.6%. The increase is not the alarming part. The alarming part is that a number this size is now simply the normal condition.
The deeper problem with the figure is what sits inside it.
Why People Are Actually Kidnapped in South Africa
Established motive as a share of kidnappings, from the SAPS Crime Registrar's analysis of 14,646 sampled cases out of 17,061 reported in 2023/24. Motive could be established in 91% of sampled cases. Source: Institute for Security Studies analysis of SAPS Crime Registrar data.
Which raises an obvious objection. Why look closely at a category this small?
Because it is moving against everything around it. In the third quarter of 2025/26 the police recorded murder down 8.7%, aggravated robbery down 11.3%, carjacking down 8.1%, business robbery down 22.5%. Kidnapping rose 6.8%.
Because these particular cases produced prosecutions, and prosecutions produce documents. Most express kidnappings leave no trial record at all, which means no testimony, no charge sheet and nothing for anyone to examine afterwards.
And because of a single sentence in a Durban judgment delivered this July, which is where this ends up.
Gauteng4. Gauteng: Volume and Location
Gauteng's share of the national figure has a dull explanation before it has an interesting one.
The dull one is arithmetic. The province accounts for 49.5% of every carjacking in the country, and roughly 80% of its kidnappings happen in the course of a hijacking or another robbery. Kidnapping in Gauteng is not usually a separate enterprise. It is a step inside an armed robbery.
The interesting explanation was delivered on 29 May 2025, at Jeppe police station, to a group of provincial legislators who had turned up unannounced.
Station management briefed the Gauteng legislature's Portfolio Committee on Community Safety on what its officers were seeing. Victims were being drawn to the Maboneng precinct from across the province through hookup apps, then kidnapped, robbed or assaulted. The committee recorded the arrest of a group operating in the area that had targeted LGBTQIA+ people specifically.
Maboneng is galleries, coffee shops and converted warehouses. For a decade it has been the standing proof that central Johannesburg could be brought back. That reputation is not incidental to what the officers described. It is an address people say yes to.
The Maboneng precinct in central Johannesburg, falling under Jeppe police station. In May 2025 station management briefed provincial legislators on victims being lured to the area from across Gauteng through hookup applications.
Denver Hostel, where the Wits student was held, sits about five kilometres away. Nobody would agree to meet a stranger there. Nobody was asked to.
Confrontation Crime Falls, Kidnapping Rises
Year-on-year change by crime category, third quarter of 2025/26 (October to December 2025). Bars to the left of the centre line are reductions; bars to the right are increases. Source: SAPS quarterly crime statistics, released March 2026.
The economics of a small ransom
There is a claim in circulation that these groups have come down-market, abandoning wealthy targets for ordinary earners and many small payments instead of a few large ones. It sounds right. The case record does not contradict it. It also does not prove it, and it is worth saying so.
What the record holds is sums. R30,000 demanded in the Wits case, R3,000 actually paid. R3,000 again in Cape Town on 11 June 2024, where the accused allegedly made a man ask a relative for money. These are the numbers in the charge sheets. They are nowhere near the figures that make headlines when a businessman is taken.
Jenni Irish-Qhobosheane of the Global Initiative Against Transnational Organized Crime describes two tiers operating at once: professional transnational syndicates working the ultra-wealthy, and local imitators far down the income scale. Children are taken in townships, she notes, where the parents do not have much money either.
That is a description of a market rather than of an organisation. The dating-app cases sit at its lower end.
5. Cape Town: The Rental Property Cases
On 14 June 2024 officers of the Hawks went into a rented house in Woodstock and found a 50-year-old man tied up inside. They arrested three men from Soweto: Mlungisi Mbuthu, 37, Bongani Zulu, 36, and Zwelakhe Ngwenya, 25.
The word that matters in that sentence is rented. No hostel, no derelict block. An ordinary house on an ordinary street, taken for a short period. A Woodstock address tells a visitor nothing. Neither does a Camps Bay one.
Nobody phoned the police. The Hawks found the house by watching odd transactions moving through the victim's bank account and following the money across several institutions.
Woodstock, immediately east of the Cape Town city centre. A rental house here was the site of the June 2024 rescue that opened the Western Cape prosecution. Charges in the same matter relate to incidents in the CBD, Camps Bay, Claremont, Durbanville and Brackenfell.
Then the arrests were reported, and other men began to come forward. By July 2024 the State had at least seven. Cases opened in Bellville were moved to Cape Town so they could be heard together. The charge sheet reached seven counts of kidnapping, extortion and robbery with aggravating circumstances, then took on cybercrime charges. That October the State said it would add at least 20 counts of money laundering.
One rescue had turned into a docket.
Two things about the case are worth setting down.
Zulu and Ngwenya already had a kidnapping and extortion matter waiting for them in Johannesburg. The prosecution says they were out on bail in it when the Cape Town offences were allegedly committed. Their lawyer, Nceba Jack, initially abandoned the Cape Town bail application so that the Gauteng warrants could be dealt with first. Investigators were separately checking whether the same pattern in KwaZulu-Natal belonged to them too.
And in October 2024 the whole thing stopped for a day because the isiZulu interpreter did not come to court. He had not been paid by the Department of Justice. A prosecution assembling twenty money laundering charges across three provinces was held up by an unpaid interpreter's fee.
What can and cannot be said about holiday rentals
The rental element is real, and the Woodstock arrest documents it. The larger claim, that groups systematically lease or take over holiday accommodation in order to attack tourists, needs handling with more care than it usually gets.
South Africa cannot measure it. Lethabo Lee Ntshudisane, who has studied the local short-term rental market, told Radio 702 in May 2026 that the police keep no category for crimes at such properties and that the platforms publish nothing. Every claim about the scale of this is therefore an inference drawn from individual cases, including the ones in this article.
The conditions, though, are not in doubt. Cape Town's short-term rental market is large and lightly regulated, currently facing a proposed by-law and a rates reclassification for properties let mainly to visitors. Remote check-in, minimal verification, high turnover. Private space for a night is easy to come by. That is an opportunity. How often anyone uses it for this, nobody can say.
6. Express Kidnapping and Mobile Banking
The police and the researchers call it express kidnapping. Hours, not weeks. The object is the victim's accounts, not their family's savings. The Institute for Security Studies puts it plainly: people are held so they can be made to draw cash at ATMs or move money through banking apps into accounts the offenders control.
Nothing about this required criminal ingenuity. The banks changed, and the crime followed. Once a person's accounts open with a fingerprint on a device in their pocket, there is little reason to break into a house.
Canada's travel advisory for South Africa is unusually blunt about it. Express kidnappings happen regularly, it tells travellers. Victims are taken from the street or from a taxi and made to withdraw cash, and some are held overnight so that a second day's limit can be used.
Read that last clause again. The length of the ordeal is set by the victim's daily transfer ceiling. The crime runs on banking rules.
Source: Institute for Security Studies analysis of SAPS Crime Registrar data; court records and reporting listed under Sources. Last refreshed: 29 July 2026.
The app does one job inside this. It removes the search. A hijacker has to be at the right intersection at the right second and deal with whoever is driving. An arranged meeting produces someone who has agreed to travel, arrives on their own, and has usually told nobody where they are, because that is the nature of the appointment.
7. Why Cases Go Unreported
Everyone who works on these cases says the same thing, and none of them can prove it. The cases that reach court, they say, are a fraction of the ones that happen. Usually there is no way to test that.
Once, there is.
When four men were sentenced in the Durban Regional Court on 15 July 2026 for luring victims through Grindr and robbing them in Ballito, Umhlanga and Durban North, the court was told that investigators had uncovered numerous additional incidents. Only three cases were prosecuted, because several other victims declined to lay charges for personal reasons.
No percentage comes out of that sentence, and it is still the best evidence in the country on how often this crime is reported. The state knew there were more victims. It could identify them. It could not act for them, because they would not come.
Their reasons are not hard to reconstruct, and they stack.
Disclosure as a second harm
To report the crime accurately, a victim must explain where they were going and why. For a man who is not out to his family, employer or church, laying a charge can mean disclosing his sexual orientation to police, prosecutors, a courtroom and potentially the press. The crime and the remedy carry the same cost.
Expectation of disbelief
The transgender activist Thabsie Mabezane, taken at gunpoint in Johannesburg after arranging a Tinder meeting, did not report the incident. Her stated reason was that she feared she would not be believed. That calculation is made before any police officer has said a word.
Case duration
The Johannesburg trial arising from a September 2023 kidnapping was still hearing prosecution witnesses in mid-2026 and has been repeatedly postponed. A victim weighing whether to come forward is weighing three years of testimony, cross-examination and delay.
Small sums, high thresholds
A victim who lost R3,000 and was released within hours may conclude that the loss does not justify the exposure. The economics that make small ransoms attractive to offenders also make them rational to absorb in silence, which is precisely why the model works.
Put those together and the arithmetic runs one way. Choose victims who are less likely to report, and the risk of prosecution per offence falls, which means more offences are possible against the same group before anyone is caught.
Which is why the targeting of LGBTQ+ people, documented by the police, by the Gauteng legislature and across successive charge sheets, is better understood as a selection criterion than as a coincidence.
None of this is a contested reading. Athlenda Mathe, the national police spokesperson, has said outright that perpetrators are targeting people through social networks including dating apps, and that syndicates going after the gay and lesbian community have been taken down. Canada tells its travellers the same thing in writing.
8. Where the Three Cases Stand
Three prosecutions mark how far the law has got. One is finished. Two are not.
Durban, concluded: convictions and 15-year terms
On Wednesday 15 July 2026 the Durban Regional Court sentenced four men for offences committed along the coast in 2023, in Ballito, Umhlanga and Durban North. Khumdulani Ngubane and Nkanyiso Ndebele were convicted on three counts of kidnapping and three of robbery each. Celimpilo Nkomo and Mzwemvelo Ngubane on one of each.
Added up, the sentences came to 57 years for Ngubane and Ndebele and 19 for Nkomo and the second Ngubane. The court ordered them served concurrently. All four walked out of the dock with 15 effective years.
The method the court heard was the same every time. Fake profiles on Grindr. An invitation to a hotel or guesthouse. An attack on arrival, then assault, restraint, and the taking of valuables along with bank cards and PINs. Some of the men guarded the victim while others went to withdraw the money.
It was not the police who first noticed them. Marshall Security, a private firm, picked up the pattern in July 2023, and its special operations team was there for the arrests later that year.
Johannesburg, active: the seven-accused trial
Seven men are on trial in the Johannesburg Magistrate's Court in connection with the September 2023 kidnapping of the Wits student: Sanele Ndlovu, Vikani Khanyeza, Sohollo Khumalo, Sphamandla Mavundla, Lungisani Mshabalala, Thulani Mazibuko and Khayelihle Zulu. They face charges of kidnapping, extortion, attempted murder and theft. All are presumed innocent.
The trial has become a study in delay. It sat in February 2026, was set down for May, was postponed, and now resumes on 12, 17 and 18 August. The defence is challenging the identification parade held during the investigation.
Representatives of the Forum for the Empowerment of Women have sat through all of it alongside the victim's family. Noxie Mtshali, speaking for the organisation, has said the proceedings are taking too long and costing too much, and that they are waiting anyway.
Cape Town, active: the three-accused matter
Mbuthu, Zulu and Ngwenya face seven counts of kidnapping, extortion and robbery with aggravating circumstances, together with cybercrime charges, with money laundering charges signalled by the State. Their attorney has indicated they will plead not guilty and has noted that the State is entitled to bring charges but must still prove them. All three are presumed innocent.
Three Prosecutions, Measured in Years
Elapsed time from the offences to the current position of each case, in months. Bars run from the month of the underlying offences to July 2026, or to sentencing where the case has concluded. Sources: court reporting by MambaOnline, EWN, IOL, News24 and SuperSport, 2023 to 2026.
- May 2023A Hillbrow detective notices a patternThe officer who would later lead the Johannesburg investigation testified that he first became aware of Grindr being used in violent crime while working another case.
- Jul 2023KwaZulu-Natal crew identifiedMarshall Security becomes aware of a group luring men to hotels and guesthouses in the Ballito and Umhlanga corridor. Arrests follow later that year.
- Sep 2023Wits student kidnapped, R30,000 demandedA multi-agency sting traces a ransom withdrawal to an ATM near Denver Hostel. The student is rescued; seven men are arrested.
- 14 Jun 2024Woodstock rescueThe Hawks find a 50-year-old man tied up in a Cape Town rental house after tracing irregular banking activity. Three Soweto men are arrested; further complainants come forward within weeks.
- Oct 2024Money laundering charges signalled, case delayedThe State indicates it will add at least 20 money laundering charges. The hearing is postponed because an interpreter is unavailable, having reported non-payment.
- 25 May 2025Olorato Mongale killedThe postgraduate student is murdered after leaving for a date. The main suspect later dies in a shootout with police. The case reframes public understanding of date-initiated violence beyond the LGBTQ+ context.
- 29 May 2025Legislators briefed on MabonengJeppe station management tells a Gauteng legislature oversight visit that victims are being lured to Maboneng from across the province through hookup apps.
- Mar 2026Gauteng establishes a specialised unitSAPS sets up a provincial unit dedicated to kidnapping and extortion. The following quarter's figures show Gauteng at 2,452 cases.
- 15 Jul 2026First convictions and sentencesThe Durban Regional Court sentences four men to effective 15-year terms. The court hears that numerous further incidents were uncovered but only three cases could be prosecuted.
- Aug 2026Johannesburg trial resumesHearings are scheduled for 12, 17 and 18 August, nearly three years after the offence.
9. What Is Not Measured
An article like this is supposed to end with a number. There isn't one, and the absence is the finding.
South Africa counts kidnappings. It does not count how they started. No field in the statistics marks a case that began on an app, no category covers offences at short-term rental properties, and the platforms release nothing.
The number most often quoted in this field, that a syndicate was taking three or four people a night, comes from a six-month internal study by a private security company that was at the time calling for Grindr to be banned in South Africa. It has never been independently verified. It is quoted anyway, including by people who should check.
What exists is a set of established points around an unmeasured space. Gauteng at 2,452 kidnappings a quarter. Four convictions in Durban. Three cases prosecuted out of numerous incidents uncovered. Ten men on trial across two provinces, presumed innocent, in matters running to 34 months. Seven complainants in one Western Cape docket that began with one. A national police service that acknowledges the pattern, a provincial legislature briefed on a specific precinct, and a foreign government that warns its travellers explicitly.
What can be said is that this crime fits the country it is happening in, and does not fit the instruments built to measure it. It needs no vehicle, no confrontation with a hardened target, no negotiation with a wealthy family. It needs a phone, a room rented for a night, and a victim with a reason to keep quiet. All three are in plentiful supply.
The July sentences prove the cases can be won and that a court will treat them as serious. The same judgment records that most of the victims in that investigation wanted no part of it.
Both things are true, and the second one will outlast the first.
10. If This Happens to You
Reporting and support in South Africa
- Emergency: 10111 from any phone, or 112 from a mobile.
- SAPS Crime Stop: 08600 10111, for anonymous information.
- OUT LGBT Well-being: counselling and support for LGBTQ+ victims of crime, including help with the reporting process where a person is concerned about disclosure.
- Forum for the Empowerment of Women: court support and advocacy, and an organisation that has attended these proceedings directly.
- Your bank's fraud line: if accounts were accessed under duress, report it as coerced access rather than ordinary fraud, and request that the transaction pattern be preserved. Banking records have opened at least one of these cases.
Questions Readers Ask
Is it true that Gauteng has 27 kidnappings a day?
Are dating apps driving South Africa's kidnapping figures?
What is an express kidnapping?
Which Cape Town areas appear in the charges?
Has anyone been convicted?
Why is the LGBTQ+ dimension emphasised?
11. Sources
Official statistics and analysis
- South African Police Service quarterly crime statistics, Q3 and Q4 of 2025/26, and annual statistics for 2014/15 and 2023/24.
- Parliamentary Research Unit, Crime Statistics Analysis: Quarter 3 (2025/26), 11 March 2026.
- Institute for Security Studies, "South Africa's armed robbery problem drives kidnapping", analysing SAPS Crime Registrar causative-factor data for 14,646 sampled cases.
- Democratic Alliance Gauteng and NovaNews reporting on the Q4 2025/26 provincial figures, May 2026.
- Global Initiative Against Transnational Organized Crime, research by Jenni Irish-Qhobosheane on extortion and kidnapping networks.
- Government of Canada travel advice and advisories for South Africa, on express kidnapping and dating-app targeting.
Court reporting
- MambaOnline, coverage of the Durban sentencing (23 July 2026) and of the Johannesburg trial (February and May 2026).
- The North Coast Courier and Marshall Security, on the KwaZulu-Natal investigation and arrests.
- EWN, IOL Cape Argus, News24 and SABC, on the Western Cape prosecution from June 2024 onward, including charge details, the centralisation of cases and the October 2024 postponement.
- TimesLIVE and SAnews, on the September 2023 Johannesburg rescue and the multi-agency sting operation.
Institutional and civil society
- Gauteng Provincial Legislature, Portfolio Committee on Community Safety media statement following the Jeppe police station oversight visit, 29 May 2025.
- OUT LGBT Well-being, South Africa LGBTIQ+ Rights Watch.
- Reuters and Thomson Reuters Foundation reporting on dating-app abductions and SAPS comment, December 2023.
- BusinessTech and Radio 702, May 2026, on the absence of any crime category for short-term rental properties.
Limitations: no official data source disaggregates kidnappings by how contact with the victim was initiated, and none records offences at short-term rental properties. Estimates of the phenomenon's scale, including those by private security firms, are extrapolations and are identified as such in the text. Case durations shown in the chart are calculated from the month of the underlying offences to July 2026 or to sentencing, and are indicative. Photography: no Wikimedia Commons image with verifiable author and licence details was confirmed at publication, and per our editorial policy no attribution has been fabricated.
This article was produced with AI assistance and reviewed and verified by humans.
View Discussion