Uganda blocked social media following 2026 general election

Last week, Ugandans were disconnected from the internet when they went to the polls for the country’s 2026 general elections. Access to the internet was shut down in Uganda between 13th to 18th January 2026, including election day (January 15th). But OONI data shows that once internet access was restored, access to specific social media and instant messaging apps – WhatsApp, Facebook, Instagram, and Twitter/X – was blocked in Uganda.

• A bit of context
• Internet shutdown during Uganda’s 2026 general election
• Social media blocks following Uganda’s 2026 general election

A bit of context

Uganda’s President, Yoweri Museveni, has been in power since 1986. And this is not the first time that access to the internet was interfered with in Uganda during elections.

We previously reported on social media blocks in Uganda during the country’s 2016 general elections and (fifth) presidential inauguration. We also reported the social media blocks and nationwide internet shutdown during Uganda’s 2021 general elections.

The internet disruptions during Uganda’s (latest) 2026 general elections seem quite similar to those during the (last) 2021 elections. In both cases, internet access was shut down completely for five days – including election day – and the shutdown coincidentally spanned 13th to 18th January in both years (2021 and 2026). Access to major social media and instant messaging platforms was blocked in the 2021 election run-up, while access to such platforms was blocked after the 2026 election.

Internet shutdown during Uganda’s 2026 general election

On 13th January 2026 – two days before Uganda’s 2026 general election – the Uganda Communications Commission (UCC) ordered a temporary shutdown of public internet access and selected mobile services during the election period. The UCC’s directive specified that all Internet Service Providers (ISPs) in Uganda had to implement the shutdown starting from 18:00 local time on 13th January 2026, and that the shutdown would be in place until instructed by the UCC to lift it. Failure to comply with the directive could result in severe sanctions, such as fines and license suspension.

Uganda’s nationwide internet connectivity shutdown is visible through several public data sources: IODA, Cloudflare Radar, and Google traffic data. All three data sources show the same pattern: a complete drop of almost all traffic originating from Uganda between 13th to 17th January 2026. This is further suggested by OONI data, which shows that almost no measurements were collected from Uganda during these days.

While OONI data is not specifically designed to provide insights into internet connectivity shutdowns, its availability depends on users’ ability to connect to the internet in order to run OONI Probe tests. As a result, the absence of OONI data may serve as a potential indicator of an internet connectivity shutdown – particularly in contexts where there is extensive, continuous, and long-term measurement coverage from a given country, and where periods of missing data align with shutdowns observed in other datasets (such as IODA).

In the chart above, we can see that while OONI measurement coverage from Uganda was relatively continuous over the last month, there is a complete absence of measurements between 14th to 16th January 2026, which correlates with the drop in internet connectivity visible in IODA, Cloudflare Radar, and Google traffic data.

Interestingly, the directive mentions that the UCC established an “exclusion list”, including a strictly defined list of systems that have been excluded from the internet shutdown and which are strictly limited to non-mobile internet services. Exemptions were granted to essential services such as healthcare, banking, and government systems.

According to the UCC directive, shutting down access to the internet during Uganda’s 2026 general elections was a “necessary measure” to “mitigate the rapid spread of online misinformation, disinformation, electoral fraud and related risks”. However, prior to the elections, the UN Human Rights Office reported that the polls were set to occur in a context marked by widespread repression and intimidation targeting political opposition members, human rights defenders, journalists, and other dissenting voices.

Social media blocks following Uganda’s 2026 general election

Internet access was restored in Uganda by 18th January 2026, several days after the country’s 2026 general election. The restoration of internet access in Uganda is visible in IODA, Cloudflare Radar, and Google traffic data.

However, even though internet connectivity was restored, OONI data shows that access to specific social media and instant messaging apps – WhatsApp, Facebook, Instagram, and Twitter/X – was blocked in Uganda. This correlates with an Uganda Communications Commission (UCC) announcement on 18th January 2026 that while internet access had been restored, access to social media and messaging apps would be restricted to “safeguard against misuse that could threaten public order”.

The following chart aggregates OONI measurement coverage from the testing of WhatsApp on 10 ASes in Uganda over the past month.

As is evident, most measurements over the past month were successful, indicating that WhatsApp was previously accessible on tested networks in Uganda. We observe a major drop in measurement coverage between 14th to 17th January 2026, correlating with the internet connectivity shutdown discussed previously. When internet access resumed on January 18th, we observed that the majority of WhatsApp measurements thereafter presented anomalies, indicating that access to WhatsApp was blocked. A per-network breakdown of the measurement coverage shows that these anomalies are mainly present on MTN Uganda (AS20294), Roke Telkom (AS37063), and Airtel Uganda (AS37075) – though other networks did not receive much measurement coverage during the same days.

However, it is important to note that a global issue affecting reachability to WhatsApp endpoints between 17th and 20th January 2026 caused our WhatsApp test to report anomalies from most vantage points worldwide. This global increase in anomalies requires further investigation. Nevertheless, OONI data indicates that access to WhatsApp was indeed blocked in Uganda: connections to web.whatsapp.com only failed in Uganda (and in a few other countries where WhatsApp is blocked), and measurements collected from 20th January onwards – after the issue was resolved and data quality was restored – continued to show consistent TCP connection failures to WhatsApp endpoints in Uganda, while succeeding elsewhere in the world.

This is demonstrated through the following charts, which aggregate OONI measurement coverage from the testing of WhatsApp Web (web.whatsapp.com) and WhatsApp’s registration service (v.whatsapp.net) on multiple networks in Uganda between 1st to 22nd January 2026.

The above charts have been generated based on the more advanced analysis of the OONI Pipeline v5, illustrating the specific errors that emerged as part of OONI Probe testing. As shown, many attempted TCP connections to WhatsApp Web (web.whatsapp.com) and the WhatsApp registration service (v.whatsapp.net) resulted in connection_refused errors from 18th January 2026 onwards, indicating that access to WhatsApp was blocked on some networks in Uganda.

Quite similarly, access to Facebook Messenger appears to have been interfered with in Uganda once internet access was restored. The following chart aggregates OONI measurement coverage from the testing of Facebook Messenger on 10 ASes in Uganda over the past month.

Unlike WhatsApp, we observe a relatively substantial presence of anomalies in the OONI Probe testing of Facebook Messenger in Uganda over the past month, prior to the internet connectivity shutdown. A per-network breakdown of the measurement coverage shows that these anomalies are mainly present on a few networks – such as MTN (AS20294) – and that Facebook Messenger was found accessible on most of the tested networks in Uganda during this period. But similarly to the testing of WhatsApp, we observe that the majority of Facebook Messenger measurements from January 18th onwards (once internet access was restored) presented anomalies. These measurements show that while DNS lookups resolved to Facebook IP addresses, attempted TCP connections to these endpoints failed – suggesting that access to Facebook Messenger was blocked.

The blocking of Facebook is further suggested looking at OONI measurements pertaining to the OONI Probe testing of www.facebook.com in Uganda over the past month.

Similarly to the testing of Facebook Messenger, the OONI Probe testing of www.facebook.com on 12 ASes in Uganda shows the presence of anomalies over the past month (suggesting that access was blocked on a few networks), and an increased spike in anomalies once internet access was restored on January 18th. Many of the anomalous measurements suggest IP blocking because TCP connections to resolved Facebook IPs failed.

Beyond WhatsApp and Facebook, OONI data also suggests that access to Instagram and Twitter/X was blocked in Uganda following the restoration of internet access, as illustrated below.

Similarly to WhatsApp, we observe that almost all measurements before the internet connectivity shutdown were successful, suggesting that Instagram and Twitter/X were accessible in Uganda over the past month. But following the restoration of internet access on January 18th, we observe a spike in anomalies, suggesting blocking attempts.

OONI measurement coverage from Uganda has always been rather limited (especially in comparison to other countries, such as Russia), presenting a limitation to the findings. However, it’s worth noting that most websites tested in Uganda over the past week were found accessible (reducing the likelihood that anomalies were caused by testing performed on unreliable networks), with the social media websites discussed in this report presenting the largest volume of anomalies. Moreover, the Uganda Communications Commission (UCC) confirmed the social media blocks, which align with the lived experience of our community on the ground.

Overall, the shift toward imposing complete internet shutdowns during elections, combined with selective blocking of major social media platforms before and after elections, suggests an effort to maximize the political impact of targeted disruptions while minimizing the economic, political, and social costs associated with prolonged connectivity shutdowns. This is further supported by the 2026 election, during which certain internet services were whitelisted and exempted from the shutdown.

If you are in Uganda, you can contribute more measurements by running OONI Probe.