Coup d’état continues to surge as the African Union and regional bodies inadvertently make coup-making profitable again

Coup d’état continues to surge as the African Union and regional bodies inadvertently make coup-making profitable again Date | 24 December 2025

Solomon Ayele Dersso, PhD
Founding Director, Amani Africa

Biruk Shewadeg, PhD
Research Consultant, Amani Africa

There is no end in sight for the new era of coups. Five years after it emerged, the resurgence of coups on the continent has continued well into the end of 2025. Following the coup in Madagascar in October, West Africa experienced, in November and December respectively, a coup in Guinea-Bissau orchestrated by the incumbent to prevent electoral defeat and an attempted coup in Benin. Since 2020, there were eleven instances of coups in nine African states. Of the eight countries that were suspended from the AU, currently seven remain under suspension.

Beyond the theory of contagion, the persistence of coups since 2020 is in part a result of an emerging policy practice on the part of the African Union (AU) and regional bodies that has lowered the costs of making coups and restored the most cherished prize of coup making, namely being recognised as legitimate leader.

The African Union’s anti-coup framework was meant to make military takeovers unprofitable. On paper, the Lomé Declaration; the Constitutive Act; the Summit Decision (Assembly/AU/4(XVI)), 2010; the Accra Declaration; and the African Charter on Democracy, Elections and Governance (ACDEG); together outlaw coups and provide for suspension and sanctions; ACDEG’s Article 25(4), in particular, bars coup leaders from contesting the elections meant to restore constitutional order. Article 25(5) adds further cost by making coup making a punishable criminal act under AU law. These provisions tell any officer contemplating a coup that even if such a coup succeeds, there is no path for auto-legitimisation through elections. In other words, Article 25(4) of ACDEG removed the most cherished prize that historically made coup-making profitable.

As military seizure of power surges in the face of major democratic backsliding on the continent, this rule is being hollowed out in practice, albeit, as discussed previously, without AU member states being regarded as having intended to dispense with the rule in Article 25(4) of ACDEG. Indeed, one of the paradoxical developments of AU’s policy response to the recent upsurge in coups is the fact that despite its declared commitment to zero tolerance to unconstitutional changes of government, in practice it has engaged not only in selective application of its policy but also created, as argued in an Ideas Indaba article, ‘a tendency of laxity in enforcing ACDEG’s Article 25(4) norm.’

In disregarding the application of this provision, the AU and regional bodies inadvertently remove the cost of staging coups, thereby making military seizure of power profitable again. In this context, instead of a bright red line, the anti-coup norm is becoming a negotiable constraint. Once that happens, as suggested previously, militaries start to read the norm not as a hard prohibition but as a risk that can be managed and bargained away.

This story of the lowering of the cost of military coups began in Chad. When Idriss Déby died in April 2021, and a Transitional Military Council (TMC) installed his son Mahamat Idriss Déby, the takeover met the criteria of an unconstitutional change of government (UCG): suspension of the constitution and military control of the transition. Instead of designating it a UCG and suspending Chad from the AU, the PSC accorded it ‘exceptional treatment’, excused in terms of Chad’s role as a frontline counter-terrorism partner, and marked a significant departure from the AU’s own rules. Perhaps more gravely, the PSC failed to enforce its decision on the non-eligibility of members of the TMC when Mahamat Idriss Déby Itno ran in the 2024 presidential election.

That moment matters because of the precedent it set. As a most recent Amani Africa Insight captures the signal this sent: coup leaders and militaries watching from elsewhere could reasonably infer that the PSC might be ‘willing not to insist on full compliance with AU norms if they opt for speeding up the ‘restoration’ of constitutional order as long as they hold elections.’ In other words, if you move quickly to elections, the AU may look the other way. That is the beginning of a classic moral hazard: the more you expect leniency after a risky act, the more attractive the act becomes.

Gabon then turns this early crack into a pattern. After Brice Oligui Nguema’s August 2023 coup removed Ali Bongo, the AU did the formal things: it recognised the UCG and suspended Gabon. That is as far as the AU went in upholding its norm on UCG. The AU looked the other way when Nguema, the very officer who led the coup and served as transitional head of state, stood as a candidate in the 12 April 2025 presidential election and won by about 90 per cent of the vote. The AU awarded Nguema by lifting Gabon’s suspension, in total disregard of the scenario ACDEG 25(4) was meant to prohibit.

While AU member states may not be considered as changing through practice the applicability of Article 25(4) of ACDEG, AU’s failure to at least reaffirm its commitment to this provision despite failure to enforce it in Gabon represented a significant normative rupture. It hollowed out the most important value of the non-eligibility rule, namely, coup makers cannot legitimise themselves through getting elected. If Chad’s case suggested you might escape suspension altogether, Gabon’s suggests that even if you are suspended, you can fairly quickly convert a coup into a continentally acknowledged electoral victory.

Guinea’s trajectory reinforces the same message. After the September 2021 coup, Guinea embarked on a drawn-out transition, with shifting timelines and growing concentration of power in the junta’s hands. Recent developments including, a constitutional referendum that allows junta members to stand in elections and extends presidential terms, are widely read as preparing the ground for coup leader Mamadi Doumbouya to run for president under rules his regime has crafted.

As pointed out previously, if the precedents set in Chad and Gabon are maintained, it would be unfeasible to bar Guinea from proceeding in the same direction. This will carry significant implications for the continuing credibility of the AU norm and for the AU’s broader engagement with other countries under political transition in respect to the application of this norm. This is where leverage shifts decisively to coup makers. When enforcement is inconsistent and sanctions are easily reversed, suspension, when it happens, becomes a temporary inconvenience rather than a real prohibition. Coup makers quickly learn that if they can stage elections on a timetable that satisfies regional fatigue with prolonged crises, they can keep both the prize and the recognition.

The result is a dangerous kind of normative slippage: ACDEG 25(4) still sits in the legal texts, and communiqués continue to reiterate the AU’s zero tolerance for UCG, including in the PSC’s 1305^th and 1306^th emergency meetings on Madagascar, but the repeated failure to enforce the non-eligibility rule in practice tells a different story. Each time the PSC validates elections where coup leaders stand, or lifts a suspension without even naming a breach of Article 25(4), it tacitly rewrites the rule. The norm survives in the legal text while erased in concrete decisions, with the policy of zero tolerance becoming more like a comforting narrative the organisation tells about itself than a binding commitment that shapes behaviour. From the perspective of militaries and political elites, the lesson is simple: the risks of staging a coup are decreasing, not only can the stigma of a coup maker be rectified by holding elections as in Gabon, but also, if you are lucky as in Chad, suspension from the AU and ineligibility for elections may all be set aside.

Reversing this normative slippage requires the AU and PSC to realign practice with the proclaimed principle. It necessitates that they close the gap between their rhetoric of upholding a policy of zero tolerance and their emerging practice of total disregard of Article 25(4) of ACDEG.  First, as argued previously, the PSC would need to publicly reaffirm Article 25(4), explicitly stating that the AU fully stands by the rule in this provision despite its non-enforceability in some cases. While this course of action may be contested in light of the Pandora’s Box opened in those precedents and allegation of a double-standard, it offers the PSC an opportunity to reassert itself as guardian of the continental normative framework and its legal instruments. Second, the relatively strong stand of the AU against coups must be coupled with a firmer stance on ‘constitutional coups’: term-limit manipulation, rigging of elections or blocking electoral processes as the collusion of political and military elites did in Guinea-Bissau recently. As long as elected incumbents can hollow out constitutional processes without attracting comparable continental pressure, military actors will find it easier to present themselves as corrective forces rather than usurpers.