The 12-Month UNSOS/AUSSOM Drawdown: Making the Benchmarks Real

The United States confirmed on 18 September 2026 that it will let the UN Support Office in Somalia (UNSOS) fund AUSSOM's logistics through a 12-month drawdown beginning January 2027, before shutting the office down entirely, a softer landing than the outright cutoff Washington threatened in July but a full year short of the 18-to-24-month transition Somalia and the African Union had requested. FIKIR Institute's earlier brief on AUSSOM's funding cliff recommended anchoring any drawdown to published, independently verified readiness benchmarks rather than a calendar alone; Washington has now set the calendar without publishing any such benchmarks. Ethiopia has already quietly handed its Burhakaba base to Somali forces and Uganda has announced a full 2027 withdrawal, meaning the drawdown clock is already running on the ground regardless of what benchmarks eventually accompany it. This brief sets out what those benchmarks should be and who should be accountable for them.

A 12-month funding drawdown has been agreed without any published readiness benchmarks attached to it, meaning troop reductions and logistics wind-down are proceeding on a fixed calendar rather than against verified Somali security force capacity. This is precisely the failure mode FIKIR Institute's prior AUSSOM brief warned against, and the risk is no longer hypothetical: Ethiopian and Ugandan drawdown actions are already underway on the ground while the 12-month clock runs without any published metric for whether Somali forces are actually ready to fill the resulting gaps.

The United States notified the African Union Commission on 1 July 2026 that it would not support any use of UN-assessed contributions to fund UNSOS beyond 31 December 2026, the date AUSSOM's mandate under Security Council Resolution 2809 expires. On 10 September, the US Mission to the AU told the Commission it would not oppose allowing the mission "appropriate time" for drawdown and liquidation, and Reuters confirmed on 18 September that UNSOS will enter a 12-month drawdown beginning January 2027 before full liquidation, rather than closing immediately. UNSOS is the logistics backbone for the roughly 11,826-strong AUSSOM force, supplying fuel, food, water, medical support and troop transport on an annual budget of approximately $500 million; a separate AU-administered troop stipend budget of $196 million for 2025 has drawn only about $48.6 million in pledges, roughly a quarter of what is required. Somalia and the AU had sought an 18-to-24-month transition; the agreed 12 months falls well short. Uganda's Chief of Defence Forces, Gen. Muhoozi Kainerugaba, said UNSOS support had already been cut by roughly a quarter and separately announced a complete Ugandan withdrawal from Somalia by 1 August 2027, though it remains unclear whether President Museveni has formally endorsed that timeline. On the ground, Ethiopian National Defence Forces handed their forward operating base in Burhakaba, a Bay region town encircled by al-Shabaab, to the Somali National Army's 60th Division on 8 August, after holding it since 2014.

The 12-month drawdown Washington has agreed to is a genuine improvement over the cliff-edge cutoff it threatened in July, but improvement in timeline alone does not address the structural problem FIKIR Institute's earlier AUSSOM brief identified: a phased exit that is phased only on a calendar, not against independently verified Somali readiness, risks repeating the pattern of drawdown outpacing the security forces meant to fill the resulting gap. Ethiopia's Burhakaba handover and Uganda's announced 2027 withdrawal are not waiting for a benchmark framework to be published; they are proceeding now, on troop-contributing countries' own timelines, while the funding drawdown clock runs in parallel without any published metric connecting the two. A 12-month period is also objectively less time than the 18 to 24 months Somalia and the AU said they needed, which means benchmarks, if they are ever published, will have less runway to be met than the parties who requested them believed was necessary in the first place. The absence of published benchmarks nine months before the drawdown clock even starts is itself a warning sign, since a genuinely benchmark-driven transition would need that framework established well in advance of January 2027, not negotiated during the drawdown itself.

• The AU, UNSOS and Somalia's federal government jointly publish specific, independently verifiable Somali National Army and police readiness benchmarks before the drawdown begins in January 2027, tying each phase of the 12-month wind-down to those benchmarks rather than calendar dates alone. Directly implements FIKIR Institute's prior recommendation, but requires the same benchmark discipline that previous regional security transitions have often lacked, and the compressed 12-month timeline leaves less room for benchmarks to be met credibly.
• Somalia and the AU seek a further extension or a benchmark-contingent pause mechanism within the 12-month window, allowing the drawdown to slow if Somali readiness metrics are not met on schedule. Preserves some flexibility against the shorter-than-requested timeline, but requires Washington to accept a conditionality it has so far resisted in favour of a fixed calendar.
• Troop-contributing countries and Somalia coordinate directly on sequencing individual handovers, following the Ethiopian Burhakaba model, rather than waiting for a unified benchmark framework from UNSOS, the AU or Washington. Allows practical, ground-level coordination to continue regardless of whether formal benchmarks are ever published, but risks the same uncoordinated, country-by-country pattern that produced Ethiopia's earlier unilateral withdrawals elsewhere.
• The drawdown proceeds on the agreed calendar without any published benchmark framework, leaving readiness assessment to each troop-contributing country's own judgement. Requires no new negotiation, but is the option FIKIR Institute's prior brief specifically warned risks a security gap outpacing Somali capacity to fill it.

• UNSOS, the AU and Somalia's federal government should publish specific, independently verifiable Somali security force readiness benchmarks before January 2027, not during the drawdown itself, since a 12-month window that starts without a benchmark framework already in place has even less time than the shorter-than-requested timeline already suggests to establish one credibly.
• Each phase of the 12-month drawdown should be explicitly tied to those published benchmarks rather than to calendar dates alone, directly implementing the recommendation FIKIR Institute's AUSSOM funding-cliff brief made before this drawdown timeline was agreed.
• Somalia and the AU should press Washington for a benchmark-contingent pause mechanism within the 12-month window, given that the agreed timeline is a full year short of what they originally requested and offers correspondingly less margin for error if readiness lags the calendar.
• Ethiopia's Burhakaba handover should be treated as a model for documenting and verifying individual base transitions, with UNSOS or the AU publishing an assessment of how that handover was conducted, since it is the closest available real-world test of whether a coordinated, verified transition is achievable under current conditions.
• International partners should establish a joint monitoring mechanism tracking both the funding drawdown's pace and Somali readiness benchmarks in parallel and publicly, reducing the chance that further unilateral troop-contributing-country withdrawals occur, as Ethiopia's and potentially Uganda's have, outside any coordinated framework.

A 12-month drawdown with no published benchmarks is a calendar, not a plan. Somalia and the AU asked for twice as much time specifically to build the kind of verified transition FIKIR Institute has recommended since the funding cliff first became public; they received half of it, with no benchmark framework yet attached. Ethiopia's and Uganda's own withdrawal timelines are not waiting for that framework to arrive, which means the window to make this drawdown benchmark-driven rather than merely calendar-driven is closing before the drawdown itself has even started.