WorldRisk

Which passport do you hold?

Risk to a traveller depends on their nationality. A country that sanctions your destination, or is sanctioned by it, changes how you are treated on arrival. Choosing here adjusts every score on the site.

We hold travel advice written for four nationalities so far. Selecting anything else shows the neutral baseline, and every page says so where it applies rather than presenting it as an assessment made for you.

Kept in your browser only. Nothing is sent anywhere, and you can change it at any time from the header.

All countries

South Africa

29Caution29 of 100, Caution band, 25 to 49good confidence(7/7 dimensions)
What this score means
Extreme75 to 100
Severe and systematic. Device search, interception and detention over online activity are documented rather than theoretical. Reconsider what you carry and what you say.
High50 to 74
Substantial and consistently evidenced measures. Device inspection, monitored networks or restricted services should be expected rather than treated as unlikely.
Caution25 to 49
Some risk, and manageable with precautions. Measures are evidenced but are narrow, inconsistently applied, or the picture is mixed.
Low0 to 24
Ordinary precautions. No systematic device search, network restriction or prosecution of visitors for online speech is evidenced.

Government travel advice is written for one citizenry, so the same country carries a different assessment depending on your passport. Change it in the header.

Showing the baseline, which weights every issuer equally. No government in our corpus publishes travel advice written for you, so this is not addressed to you specifically.

Dimensions

Each dimension is scored independently, because the traveller who cares about legal exposure for what they have posted is often not the one carrying client data. Higher means more risk.

Network threat

How hostile is the network environment to a foreign visitor, from state actors and criminals alike?

48Caution48 of 100, Caution band, 25 to 49

good confidence4 independent sourcesEvidence is ageing: contributing sources are weighted at 56% of full strength.

Why this score (5 sources)

  • ITU Global Cybersecurity Index 2024, Technical pillar: 18.57 of 20.

    International Telecommunication Union, Global Cybersecurity Index · observed 2024-12-31

  • Freedom House scored South Africa 19 of 25 on obstacles to access in Freedom on the Net, a category covering state control over internet infrastructure and gateways, regulatory and ownership control over service providers, and the independence of the regulator.

    Freedom House, Freedom on the Net, obstacles to access · observed 2025-01-01

  • Cybercriminals can compromise public Wi-Fi networks to steal personal data or credit information.

    Global Affairs Canada, Travel advice and advisories · observed 2026-10-16 · written for CAN nationals · extracted by language model

  • Investigations stretching across at least 14 countries have revealed how suspected members of rival cults, with roots in another part of the continent and carrying out online scams, are operating in South Africa.

    Daily Maverick feed · observed 2022-10-23 · extracted by language model

  • South Africa’s connected society and developed telecoms make it a prime target for cybercriminals.

    Daily Maverick feed · observed 2025-07-17 · extracted by language model

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

3 excluded sources
  • TechAfrica News, via a site scoped search, observed 2022-11-03: aged out after 4 years.
  • Daily Maverick feed, observed 2024-07-16: aged out after 2 years.
  • Kaspersky Securelist feed, observed 2023-12-01: aged out after 3 years.

Data jurisdiction

What happens to corporate or client data I carry into the jurisdiction?

32Caution32 of 100, Caution band, 25 to 49

moderate confidence1 independent source

Why this score (1 source)

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

Border device risk

How likely is my device to be searched, imaged, or seized at the border, and can I be compelled to unlock it?

28Caution28 of 100, Caution band, 25 to 49

good confidence5 independent sources

No reports were found of devices being searched, imaged or seized at this border, or of travellers being compelled to unlock them. The score below is drawn from general indicators for this country rather than from anything observed about this risk, so it describes an absence of reports rather than evidence that the risk is absent.

What the score rests on (5 indirect sources)

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

Speech and legal exposure

What legal risk attaches to what I have said online, or say while I am there?

27Caution27 of 100, Caution band, 25 to 49

good confidence7 independent sources

Why this score (9 sources)

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

2 excluded sources
  • Committee to Protect Journalists, via a site scoped search, observed 2024-04-10: aged out after 3 years.
  • The National (Papua New Guinea), via a site scoped search, observed 2021-07-01: aged out after 5 years.

Surveillance and interception

How comprehensively are communications intercepted, retained, and monitored, and is commercial spyware in use?

27Caution27 of 100, Caution band, 25 to 49

good confidence3 independent sources

Why this score (4 sources)

  • Freedom House 2026 scored ZAF 4 of 4 on freedom to express personal views on sensitive topics without fear of surveillance or retribution.

    Freedom House, Freedom in the World · observed 2026-01-01

  • V-Dem Digital Society Project 2025 scored ZAF 2.86 of 4 on how comprehensively the government monitors social media and private online communication, where higher is freer.

    V-Dem Institute, University of Gothenburg, Digital Society Project · observed 2025-12-31

  • V-Dem Digital Society Project 2025 scored ZAF 2.87 of 4 on how far privacy protection law shields communications from government access, where higher is freer. This measures what privacy law permits rather than what monitoring is practised, so it is read here as a constraint on interception rather than a measurement of it.

    V-Dem Institute, University of Gothenburg, Digital Society Project · observed 2025-12-31

  • The World Justice Project Rule of Law Index 2025 scored ZAF 0.62 of 1 on whether freedom from arbitrary interference with privacy is effectively guaranteed, a measure built from expert assessments of warrantless interception of communications.

    World Justice Project, Rule of Law Index, sub-factor 4.6 · observed 2025-01-01

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

2 excluded sources
  • Daily Maverick feed, observed 2023-10-06: aged out after 3 years.
  • MyJoyOnline, via a site scoped search, observed 2022-07-01: aged out after 4 years.

Connectivity continuity

How likely am I to lose access to the internet, or to the specific tools I depend on?

24Low24 of 100, Low band, 0 to 24

good confidence7 independent sources

Why this score (20 sources)

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

Encryption and VPN legality

Are VPNs and strong encryption lawful, and can I be required to install an application or hand over keys?

22Low22 of 100, Low band, 0 to 24

moderate confidence3 independent sources

Why this score (3 sources)

  • Some restrictions. In South Africa, all cryptography providers must be registered. Security and law enforcement agencies to make an application to a judge for a “decryption direction” which would compel a person to provide a decryption key or assistance. Failure to comply is an offence punishable by imprisonment and/or a fine.

    Global Partners Digital, World Map of Encryption Laws and Policies · observed 2019-06-01

  • V-Dem Digital Society Project 2025 scored ZAF 3.76 of 4 on whether the government regulates the internet in a way consistent with international human rights standards, where higher is freer. This does not establish whether encryption or VPN use is lawful. It measures internet regulation in general, and is used as indirect evidence because the only survey that reads encryption law directly was published in 2019.

    V-Dem Institute, University of Gothenburg, Digital Society Project · observed 2025-12-31

  • OONI recorded no confirmed blocking across 20,145 measurements of anonymity and censorship circumvention tools in ZAF over the 90 days to 2026-10-16, with 855 further results inconclusive.

    Open Observatory of Network Interference, OONI measurement data · observed 2026-10-16

Report an inaccuracy (corrections need a link to the reporting, statute or research that shows it).

Trend

02550751002026-08-162026-10-16

The overall score is up 4.4 points over 61 days, from 25.1 to 29.5. Most movement is the event overlay decaying rather than anything happening: a news signal loses half its weight every three weeks, so a score drifts back toward its structural baseline on its own. A step that holds is the one worth reading.

Check this yourself

These are not inputs to the score for South Africa. They are where to look if you want to test what is above against something we did not use.

Sources

This assessment draws on 14 sources: Auswärtiges Amt, Reise- und Sicherheitshinweise; Committee to Protect Journalists, via a site scoped search; Daily Maverick feed; Freedom House, Freedom in the World; Freedom House, Freedom on the Net, obstacles to access; Georgia Tech Internet Intelligence Lab, Internet Outage Detection and Analysis; Global Affairs Canada, Travel advice and advisories; Global Partners Digital, World Map of Encryption Laws and Policies; International Telecommunication Union, Global Cybersecurity Index; Open Observatory of Network Interference, OONI measurement data; UK Foreign, Commonwealth and Development Office, Foreign travel advice; US Department of State, Travel Advisories; V-Dem Institute, University of Gothenburg, Digital Society Project; World Justice Project, Rule of Law Index, sub-factor 4.6. Every contributing claim is quoted with its origin and date above. Model version 0.12.0, computed 2026-10-16.

If something here is wrong, say so. Corrections are handled in public, and we correct because the evidence is wrong, not because someone objects.