ANALYSIS

Two Fronts, One Army

As Abiy Ahmed forms his new government, he faces a rebel alliance in the field, rumours of disloyalty in his own barracks, and a national dialogue report that offers another way out. What he does to guard against one threat may weaken his defences against the other.

By E. Frashie

The Ethiopian Tribune · 22 September 2026

When the House of Peoples’ Representatives opens its new term in the coming weeks, Abiy Ahmed will do what every Ethiopian prime minister does at the start of a mandate: name a government. The last time he did so, in October 2021, after an election held in the middle of a war, he used the occasion to place twenty institutions under his own office, among them the National Intelligence and Security Service and the Information Network Security Agency. The cabinet he names this time will tell us more than any parliamentary speech about how he reads the danger now facing him. That danger has two faces, and they point in opposite directions.

The threat in the open

Late on Sunday 20 September, Zemene Kassie, chairman of the Amhara Fano National Movement, read a founding statement on the TPLF-controlled Tigrai TV. It announced the Ethiopian Peoples’ Forces Alliance for Survival, a coalition of seven organisations: the Amhara Fano National Movement, the Oromo Liberation Army, the Tigray People’s Liberation Front, the Ogaden National Liberation Front, the Afar Revolutionary Democratic Unity Front (Ugugumo), the Benishangul People’s Liberation Movement and the Gumuz People’s Democratic Movement. Its declared purpose is to remove what it calls the Prosperity Party’s personalist and dictatorial regime and to install an inclusive transitional government.

There is no doubt about the coalition’s reach. Its members claim constituencies in six of the country’s twelve regional states, and in Amhara and Oromia they are already at war with the state. Its cohesion is much less certain. Fano fought alongside federal forces against the TPLF in 2020–22, and Amhara and Tigray still both claim Western Tigray. Within a day of the announcement, OLA figures were describing their participation as a procedural consensus, not submission to a unified military command. The founding text commits members to resolve their differences according to the “freely expressed will” of their peoples. Applied to Wolkait, that formula would divide the alliance’s two largest partners.

Addis Ababa’s first responses were predictable. The Prime Minister’s spokesperson, Billene Seyoum, told Reuters that the groups had long been collaborating with Eritrean financial backing. Asmara has denied such accusations before, and no evidence has been published. On Monday the Ethiopian Media Authority prohibited news outlets from using Tigrai TV as a source.

The threat in the barracks

The second threat is harder to see. For some time, claims have circulated on social media that elements of the officer corps are prepared to move against the Prime Minister. Some of these claims are attributed to the exiled Oromo politician Jawar Mohammed. The Tribune has not been able to verify them, and nothing in this analysis depends on their being true. Rumours of disloyalty in the barracks shape a ruler’s behaviour whether or not they are accurate. In a system as personalised as Abiy’s, the perception of a threat is a threat in its own right.

The rumours also carry a politics of their own. In July, according to a ZeHabesha summary of a TikTok discussion, Jawar warned against coalitions formed only to topple Abiy without first agreeing on what would replace him. A transition managed by the army and a victory by an armed alliance are very different outcomes. Those who talk up the first may be positioning themselves against the second.

Nor is the idea alien to Ethiopian history. The Imperial Bodyguard moved against Haile Selassie in 1960. The Derg dismantled the monarchy from within the armed forces in 1974. Senior generals tried to remove Mengistu Haile Mariam in May 1989. And in June 2019, on the night of the killings in Bahir Dar, the Chief of the General Staff, Se’are Mekonnen, was shot dead in Addis Ababa. Every Ethiopian ruler has had reason to watch his own officers.

“Every lock he fits against his generals is a weight he hangs on his soldiers.”

The coup-proofing dilemma

Political scientists call it coup-proofing. A ruler rotates his commanders, divides the chain of command, promotes loyalists over the able, and builds parallel forces that answer to him alone. Each measure makes a coup harder, and each makes the army worse at fighting. Abiy must now do both at once: field an army strong enough to hold Amhara, Oromia and the Tigray frontier, while making sure that no part of it is strong enough, or united enough, to turn on him.

The precedent he will least want to hear is 1989. After the failed coup, Mengistu purged his senior officer corps and later had a number of its generals executed. The army that remained was loyal, and it was hollow. Within two years it collapsed before the EPRDF. The lesson is not that Abiy faces the same fate. It is that a regime which answers an open threat and a hidden one with the same instrument, a more tightly controlled army, can win the second contest and lose the first.

What the new government is likely to do

The first and most visible move will be at the top of the security state. Expect changes at the Ministry of Defence and in the general staff, and possibly further agencies brought under the Prime Minister’s office on the 2021 model. Expect, too, another round of promotions. Abiy has used mass promotion before to widen the pool of officers who owe their rank to him personally, and a larger pool of generals is a smaller risk from any one of them.

The second will be legal. The ban on Tigrai TV looks like an opening move, not a final one. The government has declared states of emergency before, nationally in November 2021 and in Amhara in August 2023, and it has designated armed movements as terrorist organisations, including the TPLF in 2021 before the designation was lifted after Pretoria. A designation of the alliance as a whole would make contact with any of its members a crime. That would reach far beyond the fighters, to intermediaries, diaspora organisers and journalists.

The third will be an effort to split the alliance. Its seams are visible, and Addis Ababa knows where they are. The ONLF can be reminded of its 2018 agreement, and the smaller Afar and Benishangul movements can be courted locally with offers of posts and budgets. Tigray’s own divisions give the government room to manoeuvre, as does the existence of Fano factions outside the AFNM. With the OLA, the government has already shown in its separate 2024 agreement with a breakaway commander that it prefers to peel off factions rather than negotiate with movements. Every member that signs a separate peace weakens the claim that the alliance speaks for Ethiopia’s peoples.

The fourth will be to externalise the threat. By naming Eritrea as the alliance’s paymaster, Addis Ababa converts a domestic insurgency into foreign aggression. That is useful for rallying nationalist opinion, and useful for binding an uneasy army to a common enemy. It is also dangerous. Combined with the government’s long-running rhetoric about access to the Red Sea, it builds a momentum towards interstate war that may prove easier to start than to stop.

The fifth will be aimed at those who spread the rumours. If talk of disloyal commanders worries the palace, it will act against the people spreading it. Jawar Mohammed was charged once before, in 2020. Renewed proceedings against him in absentia, pressure on Oromo Federalist Congress figures still inside the country, and action against journalists who amplify the claims are all plausible.

Finally, there will be the language of dialogue. The government will continue to present the National Dialogue as the peaceful route that the alliance has refused. It may repeat the gesture of 2021, when three opposition figures were brought into the cabinet, while keeping every security portfolio in trusted hands. Whether the dialogue amounts to more than language depends on what the new government does with the Commission’s report.

The Commission’s offer

The new government will not start from a blank page. The National Dialogue Commission was due to hand its report to the outgoing parliament before 22 September. It followed a 38-day conference in Addis Ababa at which more than 4,000 delegates debated over 400 agenda items under eight pillars: nation-building; government structure and electoral systems; federal cities; religious affairs; institution-building; social and economic questions; corruption and good governance; and peace-building. The full text has not yet been published, and the Commission insists that its recommendations are advisory, not binding. But the proposals reported to have emerged from it add up to the most coherent alternative yet placed before Abiy to the path of coercion described above.

The first is electoral: a move from first-past-the-post to proportional representation. Under the present system the Prosperity Party took 438 of the 486 seats decided in June. Under a proportional system, parties that now win nothing would win something, and the arithmetic of armed struggle would change. A movement that cannot win a constituency outright has little reason to lay down its arms for an election. A movement guaranteed a share of seats has more. It is also a reform that peaceful opposition parties and civic movements have pressed for years. Adopting it would reward those who chose the ballot over the gun, and show that non-violent pressure can change the rules.

The second is constitutional. Amending the 1995 constitution is the demand that unites the government’s critics and divides them at the same time. Amhara nationalists want to revise the ethnic federal settlement and Article 39, while Oromo, Somali and Tigrayan movements regard that same article as the guarantee of their existence. Like proportional representation, constitutional reform belongs to the vocabulary of peaceful protest. It was the largely non-violent demonstrations of 2015–18, the movement that brought Abiy to power, that first put the 1995 settlement openly in question. Opening the constitution through dialogue would answer the protesters in the streets, not only the fighters in the field. The alliance has avoided the question, and so, carefully, has the government. It should also be said plainly that a constitutional door opened in the name of dialogue can be walked through in either direction. Before the election, critics alleged that Abiy wanted to amend the constitution to create a presidential system. A reform that entrenched the executive would serve none of the Commission’s purposes.

The third concerns cities: metropolitan governments for the country’s major urban centres. This goes to the heart of the contest over Addis Ababa, in which Oromo nationalists assert a special constitutional interest, and over the control of fast-growing regional towns. Giving cities their own governance, instead of treating them as trophies of whichever ethnic region surrounds them, could take one of the most combustible questions off the battlefield. It would also be fiercely resisted by those who regard Oromia’s claim to the capital as settled and non-negotiable.

The fourth speaks most directly to the events of this week: that the armed movements now gathered in the Alliance for Survival be invited into talks within the framework of the Pretoria agreement. Pretoria was a bilateral deal between the federal government and the TPLF. Fano was not at the table and has never forgiven its exclusion. Extending that architecture to Amhara, Oromia and the other regions would turn a two-party ceasefire into something closer to a national settlement. The Commission has itself argued that armed groups belong in the dialogue. The question is whether a government that has just called the alliance an Eritrean project can bring itself to negotiate with it.

The fifth looks outward: negotiations with Sudan, Somalia, Eritrea and Egypt. Each relationship is a potential front. With Sudan there is the al-Fashaga border; with Mogadishu, Somaliland and sea access; with Asmara, an unfinished peace; with Cairo, the Nile and the Grand Ethiopian Renaissance Dam. Abiy told the BRICS summit this month that Ethiopia would seek sea access only through negotiation. If he means it, regional diplomacy is also domestic security, because a neighbour at peace with Addis Ababa has less reason to finance its enemies. The remaining pillars, from religious affairs and institution-building to corruption and the economy, lack the drama of these five but would matter as much to whether any settlement lasted.

“The best coup-proofing is a settlement that leaves the generals nothing to rescue the country from.”

This is where the Commission’s offer meets the barracks. Coups in Ethiopia have always been presented as rescues: of the nation from a failing emperor, of the revolution from a failing Derg. A government that opened talks with its armed opponents, reformed the electoral system and made peace with its neighbours would leave any ambitious officer without a story to tell. In effect, the Commission has handed Abiy a political alternative to coup-proofing. It is not the only option on his desk. The Reporter reported last week that Washington has lifted the arms embargo it imposed during the Tigray war, which makes the military route cheaper. Whether he takes up the report or shelves it will be decided in the same weeks as his cabinet.

The Mengistu question

Mengistu Haile Mariam hangs over this moment in two ways, and either could describe what Abiy does next. The first Mengistu went to any length in the name of Ethiopian unity. From the Derg’s early slogan, Ityopia Tikdem, “Ethiopia First”, to national call-ups, a war economy and the purge of his own officers, he justified everything as the defence of an indivisible state. A leader convinced that he alone stands between the country and disintegration can persuade himself that almost any measure is warranted. The alliance, by warning that Ethiopia is at risk of breaking apart, has handed him that argument.

The second Mengistu is the one who left. In May 1991, with his army broken and the EPRDF closing on the capital, he flew to Zimbabwe, and his departure hastened the end of the war. Most Ethiopians remember it as flight, not sacrifice, and his motives were his own. But the precedent stands: an Ethiopian ruler concluded that the state would outlast him only if he went. Abiy, too, could decide that national unity is better served by his leaving than by his staying. Such an exit could be negotiated through the very transitional process that the alliance demands and the Commission’s report points towards, and it could be made on his own terms instead of an enemy’s.

“The first path, taken too far, ends on the second — on worse terms.”

This newspaper does not predict which course he will choose, and there is no public sign that he contemplates the second. But the two paths share a premise: that the unity of Ethiopia outranks any individual, including the one in the palace. They differ in who pays. On the first, the price falls on soldiers and civilians; on the second, on one man’s ambition. The lesson of 1991 is that a ruler who follows the first path too far may find himself left with only the second, and on far worse terms than he could have had.

The hidden price

Each of these measures has a cost that will not appear in a communiqué. Emergency powers mean mass detentions, and mass detentions in Ethiopia have tended to fall along ethnic lines. Designation and media bans narrow the space for the reporting that might tell the public what is happening in its name. The drone strikes already recorded by conflict monitors in Tigray, Amhara and Oromia fall on communities, not only on combatants. And an army reorganised for loyalty is an army less able to protect civilians from anyone at all.

What to watch

The signs will come quickly. Watch who is named Minister of Defence and Chief of the General Staff, and whether more of the security apparatus moves under the Prime Minister’s office. Watch for a state-of-emergency proclamation and for any terrorist designation of the alliance. Watch for unexplained retirements, transfers or arrests of senior officers. Watch for a separate deal with any single member of the alliance. And watch the temperature of Addis Ababa’s language towards Asmara. Above all, watch whether the Commission’s report is published in full, and whether the new government’s first programme adopts any of it or merely thanks the Commission for its work.

In July, Abiy told lawmakers that even if others joined the TPLF, the government did not have “an iota of fear” for Ethiopia’s sovereignty. A government that fears nothing has no need to fortify its own barracks. The cabinet he now presents, and what he does with the report on his desk, will show which of his two fronts he fears more, and whether, in guarding against one, he leaves the other open.

E. Frashie writes long-form analysis for The Ethiopian Tribune.