A Blackout Without a Byline
On July 30, 2026, the Internet Outage Detection & Analysis project (IODA), the network-measurement platform run out of Georgia Tech's Internet Intelligence Lab, recorded a critical, multi-region connectivity disruption across Ethiopia. Oromiya, Addis Ababa, the Southern Nations, Nationalities and Peoples' Region, and Amhara all registered anomalies simultaneously, pushing the national anomaly rate to roughly 7% against a normal 1-3% baseline. As of this writing, neither Ethio Telecom nor the federal government has confirmed a cause. That silence is itself data: Ethiopia's telecom sector is state-dominated, so a disruption this synchronized across four regions is far more consistent with a deliberate network action than a coincidental infrastructure failure, even though IODA's own measurements — aggregated at the country level — cannot by themselves prove intent.
A Familiar Pattern, Not a First Offense
What makes the July 30 event newsworthy isn't its novelty — it's that Ethiopia has done this before, more than any other country on the continent. The #KeepItOn coalition, led by the digital-rights group Access Now, has documented at least 30 internet shutdowns in Ethiopia since 2016, the highest total recorded for any African country. Most have coincided with protests, communal violence, or armed conflict rather than elections specifically, though Access Now's 2026 election-monitoring page flags Ethiopia as a heightened risk this year, citing the resurgence of conflict in Tigray and Amhara, a broader crackdown on civic space, and the suspension of prominent NGOs in 2025. A single unattributed anomaly on July 30 doesn't prove another deliberate shutdown occurred. But it lands squarely inside a pattern serious enough that international monitors were already watching for it.
The Law That Exists but Isn't Used
Ethiopia is not a legal vacuum on this question — it has a statute that could govern network blocking, and the government has mostly ignored it. Article 32(5) of the Computer Crime Proclamation No. 958/2016 empowers authorities to render a computer system or data "inaccessible or blocked," but only "based on a court order." As Ethiopian legal scholar Kinfe Yilma has argued, that provision is the most defensible legal basis available for a shutdown precisely because it builds in judicial review — a check on necessity and proportionality before, not after, access is cut. In practice, Ethiopia's past shutdowns have rarely been publicly tied to a court order under this or any other statute; when the Attorney General's office has offered a justification at all, it has pointed to the 2013 law establishing the Information Network Security Agency, whose "countermeasures" clause was written for proportionate responses to cyberattacks, not blanket regional blackouts. The gap between the law that exists and the practice on the ground is the story: Ethiopia already has a mechanism that could make shutdowns accountable to a judge, and that mechanism goes largely unused.
The Case the Government Would Make — and Why It Doesn't Hold
There is a real argument for some form of network restriction during acute unrest, and it deserves to be stated plainly rather than dismissed. In a country where communal violence has repeatedly escalated through coordinated mobilization on social media, and where armed groups operate in Amhara and Oromiya, a government facing a fast-moving security threat can point to a genuine interest in slowing the spread of incitement or preventing operational coordination among combatants. Regulators in fragile, conflict-affected states are not wrong that some online content has been implicated in offline violence.
But a full network blackout is a poor tool for that problem, and international law has said so directly. The African Commission on Human and Peoples' Rights, in Resolution 362 (LIX) 2016 adopted November 4, 2016, called on member states to "guarantee, respect and protect" citizens' access to internet services and flagged shutdowns during elections as a specific concern. The UN Human Rights Council's Resolution 38/7, adopted July 5, 2018, goes further, unequivocally condemning "measures that prevent or disrupt an individual's ability to seek, receive or impart information online" and calling on states to cease them. Both instruments make the same proportionality point a blanket shutdown cannot satisfy: cutting off an entire region's access punishes millions of people who are not organizing violence, disrupts emergency services, banking, and journalism, and — because it applies indiscriminately — does nothing to distinguish a coordinating combatant from a schoolteacher checking exam results. A security response that cannot tell the two apart isn't proportionate; it's a blunt instrument dressed up as policy.
What Accountability Would Actually Look Like
The honest reform path here isn't "never restrict access" — it's "use the court-order mechanism that already exists." If Ethiopian authorities believe a specific platform or account is being used to coordinate violence, Article 32(5) gives them a route to a targeted, judicially reviewed block. That route is slower and more accountable than a switch flip at the network operator, which is exactly the point: it forces the state to show its work. Independent measurement infrastructure like IODA matters precisely because Ethiopia's telecom sector won't confirm shutdowns on its own; external detection is often the only accountability mechanism available until the government explains itself. Until Ethio Telecom or the federal government offers a cause for July 30, the responsible read is that this is one more entry in a decade-long pattern that neither domestic law nor international norms endorse — and that a country genuinely committed to proportionate security policy has a statute sitting on the books that would let it prove that commitment, if it chose to use it.