Assessment · AST-LBN-FRAME-2026-07
Recognition Now, Withdrawal Deferred: Reading the Trilateral Framework between Lebanon, Israel and the United States
The Trilateral Framework signed in Washington on 26 June 2026 hands Israel recognition on signing day and conditions any withdrawal on a disarmament the Lebanese state cannot compel. Within forty-eight hours the contest it opened had moved inside Lebanon. The most likely path over the next 60 to 90 days is a codified stalemate, no war and no peace, the occupation hardening behind it.
The call, up front
Key Judgments
A confidence tag states how strongly the evidence supports a judgment. It is not a probability.
Recognition banked, withdrawal removed. Paragraph 5 lets Israel declare when the threat that justifies its presence has ended, and the leaked annex makes redeployment "phased, conditions-based, progressive," set by field assessment on no calendar. Israel took direct recognition and a state-monopoly and disarmament process on signing day and is bound to no withdrawal, which converts a military occupation into an open-ended hold under a peace label.
Bound in politics, unbound in law. The ambassador's signature authenticates the text but concludes nothing under Article 52, which vests treaty conclusion in the President and Prime Minister subject to Council of Ministers approval. Under Article 65 any cabinet adoption of the Framework or its annex needs a two-thirds majority the Amal and Hezbollah ministers can deny, paragraph 13 forfeits Lebanon's recourse in international fora and equates the release of unlawfully held Lebanese with the return of remains, and no authenticated Arabic text exists, a defect under Article 11. The instrument therefore binds Lebanon politically and, in parts, is arguably void, with constitutional review biting only at a cabinet adoption or the comprehensive treaty paragraph 12 promises.
The conflict renamed. Paragraphs 4 and 5 route the withdrawal that Resolutions 425 and 1701 impose on Israel onto a Lebanese act instead, the verified disarmament of Hezbollah, and the text names no other file, no borders clause, no prisoners obligation, no airspace provision, no right of return. It erases every prior reference point, from Resolutions 1701, 1559, and 425 to General Assembly Resolution 194, the 1949 Armistice, and the 2002 Arab Peace Initiative, and asks of Israel only a self-defined declaration of "no territorial ambitions in Lebanon," recasting a state-to-state conflict over Lebanese land as one party's arms.
The decisive fracture is internal. Speaker Berri rejects the text as dictates "ten times worse than the 17 May 1983 agreement" and Hezbollah's bloc calls it void, while President Aoun owns a signed instrument he cannot ratify, so the contest has moved from the front line into ratification and into the question of the state's legitimacy to sign at all. An observable public discourse converts the demand on Hezbollah into a loyalty test of the Shia community, opposition is not confined to the Shia parties, and as of the cutoff the fracture had not yet reached a cabinet or parliamentary vote.
A commitment the state cannot deliver. No Lebanese financial-control instrument implementing paragraph 11 exists, disarmament by force sits beyond the Lebanese Army's capacity and consent, and the one path that might have squared the commitment, integrating Hezbollah members into the Army, a proposal Aoun himself carried through 2025, is the single option the text explicitly rules out. The state has thus committed to a service it cannot render, verified Hezbollah disarmament on a timeline Israel controls.
No war, no peace. Strikes, demolitions, and more than 700,000 displaced at the June peak persist after signing, and reconstruction and return are now conditioned on a disarmament benchmark that forecloses the very return the text promised. The Framework most likely entrenches the occupation and displacement it promised to reverse, leaving neither war nor peace.
The web edition
What the instrument does
The Trilateral Framework signed in Washington on 26 June 2026 hands Israel recognition on signing day and conditions any withdrawal on a disarmament the Lebanese state cannot compel. Within forty-eight hours the contest it opened had moved inside Lebanon, into whether the state may ratify it at all and into a loyalty test pressed on one community.
To make the trade work, the Framework had to rename the conflict: a dispute older than Hezbollah, over occupation, borders, prisoners, and violations of Lebanese airspace and waters, becomes a problem of one Lebanese party's arms. That redefinition erases every prior reference point from Resolution 1701 to the Arab Peace Initiative and routes nearly every operative obligation onto the Lebanese state, while asking of Israel only a self-defined assurance. It is the hinge the rest of the instrument turns on.
As signed, the Framework binds Lebanon politically but carries no legal force, and its defects leave it contested and in parts arguably void. The ambassador's signature concluded nothing, the constitutional chain of Articles 52 and 65 is uncleared, the Amal and Hezbollah ministers hold a cabinet blocking third, and no authenticated Arabic text exists. Nor was the asymmetry forced on Lebanon: the concessions were made at a table across five rounds, not extracted on a battlefield, and the one asset that might have moved Israel, recognition, was given first and unpriced.
The instrument belongs to a recognizable family, the framework agreements from Camp David to Oslo to the 2003 Roadmap that bank recognition up front and defer the hard file into talks that bury it. Its nearest Lebanese precedent, the 17 May 1983 agreement, reached further through the constitutional process than this text has and still collapsed within a year, which is the measure of the "irreversible" the document claims and cannot secure.
Read as reservation points, the two floors that matter miss each other. Israel will not date a withdrawal it does not control, and Hezbollah will not disarm while the occupation stands. The stalemate is derived before any forecast of how the actors will behave.
The next 60 to 90 days
Scenarios
These likelihood badges are subjective probabilities over the 60 to 90 day horizon. Because the six paths are not mutually exclusive, the bands are standalone and are not constrained to sum to one. The ranking rests on the reservation-point ledger: the stalemate dominates because it alone requires no actor to cross its floor.
Codified stalemate: promise without delivery. The Framework holds as signed, the pilot zones stall or crawl, no withdrawal timetable materializes, the ceasefire persists intermittently under continued strikes, the occupation hardens behind the belt, and displacement settles toward permanence. The path dominates because it requires no party to concede its core and matches every actor's stated posture, and because each actor can blame the stall on another indefinitely.
Internal rupture over ratification. The file is forced to a Council of Ministers vote, the Article 65 two-thirds test makes the blocking third show itself, and the contest escalates from veto to legitimacy crisis, a government paralyzed or fractured, with the loyalty-test discourse hardening into sustained intra-Lebanese confrontation.
An internal settlement on the arms. The state and Hezbollah reach an accommodation that neither the Framework's terms nor its verifier would recognize, a Lebanese formula on the weapons that trades the text's benchmark for a domestic one.
The remaining paths, renewed US pressure, regional absorption of the file, and a collapse at the front, sit below these in likelihood. The full assessment carries all six with their bands, the indicators to monitor for each, and the assumptions whose failure would change the reading.
How this was read
The method, so it can be checked
The assessment reads the instrument across its legal, political, and operational dimensions. Judgments carry confidence tags on the strength of the evidence. Scenarios carry likelihood bands, and the two are kept strictly apart. The central judgment is derived from the interests and reservation points of the actors, not asserted: the ledger is set out in Part Three so the derivation can be checked. Annex-dependent judgments are flagged with their provenance.