Larak Strike Reopens Direct US-Iran Exchanges and Restricts Hormuz Shipping

Bottom Line

There is a 55-70% chance of at least one additional direct US-Iran exchange involving Iranian, US, Jordanian, Emirati or commercial targets between 2 and 30 September 2026, with moderate confidence. The Strait of Hormuz is likely to remain operationally restricted, defined as fewer than 14 visible commodity-vessel transits on a seven-day average or continued major-carrier avoidance. Normal crude and LNG movement is very unlikely to return by 30 September, also with moderate confidence.

The Larak and Jordan sequence is supported by CENTCOM, the Jordan Armed Forces, the UAE Ministry of Defence, wire services and major media. Iranian damage claims and the alleged strike on Al Minhad Air Base remain disputed. The evidence supports severe maritime disruption, but not a verified physical closure of Hormuz.

Key Developments

Reported developments

• On 30 August, CENTCOM reported a strike on two Iranian launchers on Larak Island. CENTCOM said the launchers were being prepared for sea-mine operations in the Strait of Hormuz. This is primary US operational reporting and is self-interested on the question of US intent, but the strike itself is corroborated by multiple major-media and wire-service accounts.
• The IRGC reported missile and drone attacks towards King Hussein and Al-Azraq air bases in Jordan. The Jordan Armed Forces reported intercepting eight missiles. The existence of Iranian launches and Jordanian interceptions is better supported than Iranian claims of damage.
• The UAE Ministry of Defence reported intercepting one Iranian drone over UAE territorial waters and denied that Al Minhad Air Base was hit. The Emirati account is primary national reporting; the competing Iranian claim is self-interested and not independently verified.
• Kpler reporting recorded five visible commodity vessels per day over one weekend and 14 vessels on a reported Friday. These figures measure different periods and do not establish a seven-day average, a pre-crisis baseline or total traffic. Some vessels reportedly switched off AIS.
• The Ukrainian General Staff reported 203 combat engagements on 25 August, including 36 Russian assaults near Pokrovsk and 24 near Kostiantynivka. Russian and Ukrainian territorial claims remain opposed.
• Ukraine reported more than 550 Russian drones and loitering munitions during the 28-29 August strike series, including Shahed, Gerbera, Parodiya and S8000 Banderol systems, alongside Onyx and Iskander-M missiles. The strike volume is supported by Ukrainian official reporting and major media; exact launch totals remain party-reported.
• OCHA, WHO, UNICEF and WFP reported continued Israeli attacks and severe humanitarian conditions in Gaza. The reported indicators include 50% of families running out of food at least once in the previous month, about 90% of civilian infrastructure heavily damaged or destroyed and more than two million displaced people.
• UNIFIL reported increased logistical movements, demolitions and projectile launches in southern Lebanon. Hezbollah drone attacks and Israeli strikes continue, but the supplied record does not establish Iranian operational direction of these incidents.
• Nigerien authorities reported containing an armed mutiny involving Air Base 101, Diori Hamani International Airport and areas near the Presidential Palace between the night of 28 August and 29 August. Russia's ambassador reported assistance from Africa Corps. The precise Russian role, casualty total and control timeline remain unresolved.
• The SAF and RSF continued combat around Bara, El Obeid, Darfur and Blue Nile. A reported strike near the Chad-Sudan border remains disputed and is not independently corroborated in the supplied record.
• The gang assault on Kenscoff during the night of 23-24 August killed at least 47 people, displaced more than 2,600 and involved more than 50 abductions, according to UN, BINUH, wire-service and local reporting. The event is strongly corroborated, although individual casualty and hostage figures vary.
• The PLA Southern Theater Command and China Coast Guard continued pressure around Scarborough Shoal and Second Thomas Shoal. The record does not corroborate a new dated PLA exercise or armed China-Taiwan incident during the reporting window.

Source quality and auditability

The 60-brief package contains 196 reports assessed as reliable, 269 as usually reliable, 66 as fairly reliable, 52 as not usually reliable, 29 as unreliable and three as not assessable. These are report counts, not independent-source counts. Repeated claims across briefs therefore do not necessarily constitute independent corroboration.

Primary operational reporting from CENTCOM, the Jordan Armed Forces, the UAE Ministry of Defence, Ukrainian and Russian military authorities, Nigerien authorities, UN agencies, UNIFIL, MARAD and NATO is useful for statements of what each organisation claims to have done or observed. It is less reliable for enemy losses, damage attribution, intent and battlefield control. Kpler and AIS data are useful for observed vessel behaviour but differ in denominator, time period and visibility. NASA FIRMS records heat, not the cause of the heat, and the supplied Taiwan and Bolivia samples are insufficient to assess military or political activity. Think-tank, blog, social-media and local-media reporting is retained where it provides unique detail, but forward judgments based primarily on those streams carry low confidence.

Analysis

What changed since 1 July

US-Iran sequence and competing explanations

The base assessment is that the Larak strike and Jordan response have increased the chance of another direct exchange, not that they prove a deliberate US-Iran decision for unrestricted war. The deliberate-escalation hypothesis is supported by the reported preparation of sea mines, the IRGC's stated retaliation and US statements about further pressure. The limited-signalling hypothesis remains viable because the Larak strike was geographically narrow and Jordanian authorities reported that the missiles were intercepted without major damage. The inadvertent-escalation hypothesis remains viable because disputed damage claims, air-defence activity and maritime encounters create conditions in which a local event can produce a wider response.

Discriminating evidence within 14 days will be another US strike on Iranian launchers or a confirmed Iranian attack causing casualties. A public US-Iran channel that produces an operating maritime rule without a further strike would favour limited signalling and risk management. Repeated attacks on US bases, commercial vessels or Gulf infrastructure would favour deliberate escalation.

Hormuz measurement and competing explanations

The 80-95% judgment that commercial movement remains restricted rests on convergent evidence from low visible traffic, carrier caution, reported maritime incidents, higher insurance costs and competing official accounts. The 20-45% probability of a near-total halt is lower because vessels continue to transit and the collection does not prove that Iran controls every route. Low traffic can reflect Iranian selective control, voluntary commercial avoidance, AIS undercounting or all three. A seven-day count with a declared denominator is required before assigning a percentage change from normal traffic.

Russia-Ukraine

Russian and Ukrainian reporting supports continued high operational tempo, but not a decisive change in the land campaign. Russian pressure is concentrated around Pokrovsk, Kostiantynivka, Sloviansk and Lyman. Ukrainian forces retain local counterattack capacity near Dobropillia, Kupiansk and southern Donetsk. The conflicting Russian claim of control over Kostiantynivka and Ukrainian and independent mapping that place the city under Ukrainian control prevent a high-confidence territorial judgment.

The Russian and Ukrainian long-range campaigns are more firmly established than their strategic effects. Russian strikes are very likely to continue, at 80-95%, because the 28-29 August operation involved a large reported number of systems and Kyiv experienced repeated attacks. Ukrainian strikes are likely to continue, at 65-80%, because repeated attacks on refineries, airfields, warehouses and maritime logistics are reported by Ukrainian authorities and Russian regional authorities. The stronger claim that Ukraine has removed 25-30% of Russian refining capacity or caused a historic fuel crisis remains insufficiently corroborated.

Gaza and southern Lebanon

Gaza and southern Lebanon are assessed separately. Israeli attacks in Gaza are very likely to continue, at 80-95%, because OCHA, UNICEF, WHO and WFP report ongoing attacks and no verified ceasefire enforcement mechanism. Severe humanitarian conditions are very likely to persist, at 80-95%, because food, water, infrastructure, displacement and medical-access indicators are corroborated by multiple UN agencies.

Hezbollah-Israel exchanges in southern Lebanon are likely to continue, at 55-70%, because UNIFIL and local reporting describe repeated projectile launches, drone activity, demolitions and Israeli strikes. Direct Iranian operational involvement in Lebanon is not established. A direct Iran-linked expansion is roughly even chance, at 40-60%, with low confidence. Repeated attacks causing Israeli fatalities, sustained Israeli operations in Beirut's southern suburbs or independently verified Iranian support to the Lebanon front would revise the estimate upwards.

Niger

The mutiny is almost certainly a real armed disturbance, at 80-95%, but the responsible units, leadership and political objective remain unresolved. An isolated disciplinary revolt would be supported by rapid demobilisation, limited arrests and the absence of command changes. A factional challenge would be supported by senior officer dismissals, coordinated arrests, public criticism of Tchiani or further unrest at military sites. An attempted coup would be supported by evidence of a named leadership group, a political programme, control of additional state institutions or an effort to replace the CNSP.

The current collection supports containment of the immediate disturbance but does not support a probability judgment that a second mutiny is unlikely. The collection is sparse, relies partly on official and aligned reporting, and contains conflicting accounts of the airport perimeter, casualty totals and foreign intervention. The more defensible judgment is that Tchiani likely retained control of Niamey by 30 September, at 80-95%, while the depth of the cohesion problem remains low confidence.

Islamic State and JNIM pressure is separate from the mutiny. Continued attacks on Nigerien security forces are likely, at 55-70%, but the supplied material does not demonstrate a region-wide jihadist surge or a link between the Niamey attackers and either organisation.

Secondary theatres

The SAF and RSF are very likely to remain in combat through 30 September, at 80-95%, with moderate confidence. The direction of humanitarian deterioration is more reliable than individual displacement totals because the supplied figures use different dates and definitions.

Haiti's Viv Ansanm and aligned gangs are likely to continue testing state control around Port-au-Prince, at 55-70%, with moderate confidence. The Gang Suppression Force's 1,083 deployed personnel as at 8 July, compared with an authorised ceiling of 5,550, indicates a capacity gap but does not predict the outcome of individual operations.

Venezuela's governing order is likely to remain contested through 30 November, at 55-70%, with moderate confidence. The announced oil framework involving Delcy Rodríguez and US-linked actors is likely to advance as a political framework, at 55-70%, but its legal terms, duration and control arrangements remain disputed. This theater does not alter the Hormuz judgment.

The Houthi threat to Saudi-linked shipping is likely to remain elevated, at 55-70%, while Bab-el-Mandeb traffic is likely to remain constrained rather than fully closed, also at 55-70%. The absence of an independent current traffic count lowers confidence. Russian-linked cyber, sabotage and influence activity in Europe is very likely to continue, at 80-95%, while a direct Russian conventional attack on NATO is unlikely, at 20-45%.

Current Taiwan collection does not establish a new PLA exercise or armed China-Taiwan incident. That is a collection finding, not evidence of de-escalation. Chinese coercive pressure in the South China Sea is more strongly evidenced and is likely to continue, at 55-70%, while an armed China-Philippines exchange is very unlikely, at 5-20%.

Analytic implications for stakeholders

Chronology annex

Indicators & Warnings

• Additional direct US-Iran exchange: Watch for another CENTCOM strike on an IRGC maritime-denial site, Iranian launches towards Jordan, Qatar, the UAE or Saudi Arabia, or confirmed damage at a US or commercial facility. Baseline: one Larak strike and one reported Jordan response by 31 August. Threshold: one independently corroborated additional exchange. Horizon: 0-14 days. Owner: CENTCOM and regional air-defence authorities.
• Hormuz traffic: Watch for Kpler or AIS-derived seven-day commodity-vessel counts, separated from all-vessel counts, alongside carrier voyage notices. Baseline: five visible commodity vessels per day over one reported weekend and 14 on one reported Friday. Threshold: at least 14 per day for seven days with major-carrier participation supports de-escalation; fewer than three per day for seven days supports near-total interruption. Horizon: 7-30 days. Owner: maritime security analysts and shipowners.
• Iran-Oman procedure: Watch for a signed transit procedure naming fees, inspections, communications and eligible vessels, confirmed by Oman, Iran and an independent carrier. Threshold: seven consecutive days of compliant transits. Horizon: 7-30 days. Owner: Oman, IMO and maritime operators.
• Commercial maritime attack: Watch for a second confirmed tanker attack within 14 days, a mine detonation or a boarding documented by crew and authorities. Threshold: two independently corroborated incidents or one mine detonation. Horizon: 0-14 days. Owner: UKMTO and commercial maritime security centres.
• Ukraine ground movement: Watch for three independent geolocated observations of Russian entry into Kostiantynivka or a sustained approach to Sloviansk, or repeated Ukrainian advances that threaten routes near Dobropillia or Kupiansk. Horizon: 14-90 days. Owner: OSINT battlefield assessment teams.
• Gaza access: Watch for a seven-day halt at Kerem Shalom, further attacks on water or medical facilities, or two consecutive UN reporting periods showing worsening food access. Horizon: 0-30 days. Owner: OCHA, WFP, WHO and UNICEF.
• Lebanon widening: Watch for two Hezbollah attacks causing Israeli fatalities in 30 days, repeated Israeli operations in Beirut's southern suburbs or verified Iranian operational support to Hezbollah. Horizon: 0-90 days. Owner: UNIFIL and regional security monitors.
• Niger cohesion: Watch for a second disturbance at Base 101 or the Presidential Palace, mass arrests or dismissals involving at least five senior officers, or verified Africa Corps armoured movement into Niamey. Horizon: 0-30 days. Owner: regional political-security analysts.
• Sahel jihadist exploitation: Watch for two Islamic State Sahel or JNIM attacks on Diori Hamani International Airport, Base 101, Niamey road links or redeployed Nigerien units within 30 days. Horizon: 14-60 days. Owner: Sahel security-monitoring teams.
• Sudan-Chad escalation: Watch for two independently corroborated cross-border kinetic incidents or a seven-day Chad-Sudan border closure. Horizon: 0-45 days. Owner: UN and regional conflict monitors.
• Haiti election access: Watch for one additional commune seized for seven days or one confirmed attack on an election facility. Horizon: 30-90 days. Owner: BINUH, Haitian election authorities and humanitarian monitors.
• European hybrid activity: Watch for forensic attribution of the WB Electronics or Leipzig incidents, another fire at a named defence plant, a destructive cyberattack on weapons logistics or a Russian drone incursion causing casualties on NATO territory. Horizon: 14-90 days. Owner: NATO and national security services.
• South China Sea coercion: Watch for Chinese Coast Guard boarding, ramming or water-cannon use against Philippine vessels, or a seven-day Second Thomas Shoal resupply interruption. Horizon: 14-90 days. Owner: Philippine Coast Guard, China Coast Guard and independent maritime monitors.
• Taiwan collection gap: Watch for dated PLA Eastern Theater Command activity, Taiwan Ministry of National Defence alerts, commercial satellite evidence or sustained AIS coverage around Taiwan. Threshold: two independent sources confirming a new exercise or armed incident. Horizon: 0-30 days. Owner: Taiwan and regional collection teams.
• Bolivia verification: Watch for dated Bolivian transport data, geolocated protest imagery, statements by named trade unions and independent local reporting. Threshold: two independent sources confirming nationwide disruption. Horizon: 0-30 days. Owner: regional political-monitoring teams.

Alternatives

Managed continuation, 55%

This base case is likely, at 55-70%. It applies if no additional exchange causes confirmed fatalities or major damage on new sovereign territory, no mine detonation stops commercial movement and no verified maritime arrangement produces seven consecutive days of sustained carrier participation. Selective Hormuz transits continue, major crude and LNG movements remain below normal, Russia and Ukraine continue reciprocal strikes without a decisive breakthrough, Israeli operations continue in Gaza, Hezbollah-Israel exchanges remain localised to southern Lebanon and the SAF and RSF remain in combat.

Hormuz-centred escalation, 25%

This outcome is unlikely, at 20-45%, but consequential. It applies if there is one additional direct US-Iran exchange causing confirmed casualties or major damage, two independently corroborated commercial attacks within 14 days, a confirmed mine detonation or a seven-day visible commodity-vessel average below three. US strikes on additional IRGC assets, wider Iranian attacks and further Houthi pressure on Bab-el-Mandeb would raise energy, insurance and food-import costs. The main alternative explanation is that one additional exchange remains a limited warning and does not produce sustained maritime disruption.

Verified maritime de-escalation, 20%

This outcome is unlikely, at 20-45%. It applies if there is no additional direct exchange after 1 September, Oman and Iran publish a signed transit procedure, at least one US-linked authority acknowledges it and visible commodity traffic averages at least 14 vessels per day for seven consecutive days with major carriers participating. The arrangement would reduce immediate shipping and insurance pressure without resolving sanctions, Iran's nuclear programme, compensation disputes or US force posture. Low traffic caused only by voluntary avoidance or AIS gaps would not qualify.

Outlook

From 2 to 30 September, the principal forecast is continued US-Iran pressure with a 55-70% chance of another direct exchange. Hormuz commercial traffic is likely to remain restricted, at 80-95%, while a near-total halt is unlikely, at 20-45%. A verified maritime arrangement is unlikely, at 20-45%. Jordan, the UAE, Saudi Arabia, Qatar, Oman, Pakistan, shipowners, insurers, LNG buyers and European energy markets face the clearest immediate consequences.

From 1 October to 30 November, Russia and Ukraine are likely to sustain reciprocal long-range strikes and local ground pressure without a decisive breakthrough. Israeli strikes in Gaza are very likely to continue, and severe humanitarian conditions are very likely to persist. Hezbollah-Israel exchanges are likely to continue in southern Lebanon. The SAF and RSF are very likely to remain in combat through 30 September, and continued fighting after that date is likely. Niger's government is very likely to retain control of Niamey, but the depth of military factionalism remains unresolved. Houthi pressure, Russian-linked hybrid activity in Europe and Chinese Coast Guard coercion around Philippine-held positions are likely to continue.

The Taiwan and Bolivia judgments remain limited by collection gaps. The absence of a corroborated Taiwan incident does not demonstrate de-escalation, and the Bolivia package does not substantiate the alleged current protest crisis. These judgments should be revised only when dated, geographically relevant and independently corroborated reporting becomes available.