Hormuz Stays Severely Constrained as Houthi Partial Control and Western Ukraine Strikes Raise Spillover Risk

Bottom Line

US-Iran maritime confrontation is very likely to continue from 14 through 28 September 2026, with moderate confidence. UKMTO reported an unidentified projectile striking a Panama-flagged tanker in the Strait of Hormuz on 13 September, while Iranian authorities separately reported a vessel strike near Qeshm Island. Commercial traffic is very likely to remain severely constrained, but continued movement and inconsistent traffic counts do not establish a complete closure.

Ansar Allah likely has a presence and partial tactical control at Mokha, Dhubab and Mayyun, also known as Perim, with moderate confidence. This assessment does not establish uncontested administrative control of those sites, the full Yemeni Red Sea coast or Bab al-Mandab. Russia's 13 September air campaign struck western Ukrainian transport infrastructure, including a train locomotive roughly 2 kilometres from Poland, increasing miscalculation concern without establishing an attack on Poland or NATO territory.

Key Developments

Strait of Hormuz

Reported: UKMTO, an official maritime safety authority with direct proximity to incident reporting, recorded a projectile strike on a Panama-flagged tanker on 13 September. The vessel caught fire and its crew was evacuated. UKMTO did not identify the attacker. Iranian state media and the Qeshm governor separately reported that an Iranian commercial vessel was struck near Qeshm Island, killing one person and injuring three or four. US maritime co-ordination continues to limit protected tanker movements to two daily windows.

Assessment: US-Iran maritime confrontation is very likely to continue through 28 September, with moderate confidence. The occurrence of the UKMTO incident has high confidence. Attacker attribution for both incidents has low confidence. Hormuz traffic is very likely to remain below the pre-conflict baseline, but a complete closure is unlikely because some commercial and protected movement continues.

Bab al-Mandab and Saudi infrastructure

Reported: Reuters, Associated Press and Yemeni sources described Ansar Allah advances or positions at Mokha, Dhubab and Mayyun. Yahya Saree said non-Saudi shipping remained safe, while other Houthi statements threatened Saudi-linked vessels. Saudi Arabia shut the East-West Pipeline after reported drone attacks. Iraqi reporting placed the launch origin inside Iraq, particularly Maysan, but did not identify the operators. Satellite imagery and other reporting indicate fire or damage near pumping stations, while Saudi statements also describe the shutdown as precautionary.

Assessment: Ansar Allah likely has partial tactical control at Mokha, Dhubab and Mayyun through 28 September, with moderate confidence. Bab al-Mandab traffic is likely to remain selectively constrained, especially for Saudi-linked shipping, but full closure is unlikely. The pipeline outage is likely attack-related, with moderate confidence, although a precautionary shutdown after limited damage remains plausible.

Ukraine and Poland

Reported: Ukraine's Air Force reported 453 aerial attack means during the 13 September reporting window. Other Ukrainian accounts cited 323 drones between 06:30 and 18:00 and nearly 500 drones since the morning. The different figures cover different windows and categories. Impacts were reported at Odesa, Kyiv Oblast, Boryspil, Ternopil, Odesa and Chornomorsk ports, and railway infrastructure near the Yahodyn-Dorohusk crossing. A Kyiv-Warsaw passenger-train locomotive was hit roughly 2 kilometres from Poland. A gas station and trucks were struck less than 1 kilometre from the crossing. Poland reported no airspace violation and no casualties.

Assessment: Russia is very likely to continue strikes against western Ukrainian transport, port and logistics infrastructure through 28 September, with high confidence in continued strike activity and moderate confidence in specific military targeting. It is unlikely, with moderate confidence, that the passenger train itself was the intended target. An intentional Russian attack on Poland or NATO territory remains very unlikely. The episode raises qualitative concern about accidental spillover, but the prior briefing did not provide a comparable probability for that specific risk.

Secondary watchlist

Reported and assessed: SAF and RSF drone activity is very likely to continue through 14 October, and Sudanese humanitarian conditions are very likely to worsen. General Abdourahmane Tchiani's junta is very likely to retain Niamey through 14 December, but further military tension inside Niger is likely. China Coast Guard patrols east of Taiwan are very likely to continue, while an active blockade by 28 September remains very unlikely. Israeli military activity in southern Lebanon and Gaza is very likely to persist. Russian-linked hybrid activity in Europe is likely to continue through 14 December, although individual attribution remains uneven.

The source base is strong for broad incident and humanitarian trends. The packet contains 589 underlying records, including 418 in the two strongest reliability bands, 84 usable with reservations and 87 weak or unassessable. UKMTO, UN agencies, OCHA and IOM have strong proximity to maritime safety or humanitarian conditions. Official Iranian, Houthi, Russian, Ukrainian, Saudi and Yemeni claims provide access to each actor's own actions and policies, but also carry incentives to shape attribution. Commercial imagery confirms physical effects more reliably than perpetrators. Nigerian, Bolivian and Venezuelan political reporting, Houthi control claims, the Qeshm attribution and several European hybrid cases remain thin or contested.

Analysis

Maritime disruption is severe without verified closure

The evidence supports a hazardous and heavily restricted operating environment in both chokepoints, not a proven physical closure. The US two-window protection system, Iranian reopening conditions, Houthi exemptions for some shipping categories, tanker avoidance and increased anchoring indicate that commercial operators are adjusting behaviour. AIS-visible vessels, US-facilitated movements and crude-flow estimates measure different populations, so they cannot be combined into a single transit count.

The prior assessment is unchanged: Hormuz traffic is very likely to remain far below its pre-conflict level, and a complete closure is unlikely. The 13 September UKMTO incident strengthens the evidence for immediate commercial danger, but it does not resolve the identity of the attacker or prove a unified US-Iran operational sequence.

A limited Iran-Oman or Gulf shipping arrangement is roughly an even chance by 28 September. It is unlikely to produce durable regional de-escalation because Iran continues to condition reopening on US concessions, US protection measures remain in place and no enforcement mechanism has been reported.

Houthi territorial control and maritime effect are separate judgments

The strongest current evidence supports Ansar Allah presence and partial tactical control at Mokha, Dhubab and Mayyun. It does not establish uncontested administrative control. The Yemeni government reports counter-attacks and withdrawals, the Houthi military's own statements are designed to shape commercial behaviour, and the packet lacks persistent independent imagery confirming a stable control map.

The maritime effect is more limited than the territorial rhetoric. Bab al-Mandab traffic is likely to remain selectively constrained, especially for Saudi-linked vessels, because Ansar Allah has access to the Yemeni approaches and has declared a Saudi-specific threat. Continued passage by non-Saudi vessels and the absence of verified all-vessel interdiction argue against a complete closure.

Western Ukrainian strikes raise miscalculation risk without proving Polish targeting

The 13 September attack should be treated as a defined reporting window, not a single reconciled total. Ukraine reported 453 aerial attack means, while other accounts used narrower time windows and different categories. The occurrence of impacts across Odesa, Kyiv Oblast, ports, railways and the Yahodyn-Dorohusk area is better supported than the exact launch or interception count.

Russia's Defence Ministry said it targeted western Ukrainian infrastructure used for military cargo from Europe. That statement, the affected railway and port sites, and the proximity to Poland support a likely deliberate campaign against Ukrainian transport infrastructure. They do not prove that the passenger train was the intended target. Poland's report of no airspace violation and no casualties supports the assessment that the incident remained contained within Ukraine.

Separate regional fronts should not be treated as one command structure

The US-Iran maritime confrontation, Israel-Hezbollah activity, Israel-Gaza operations, the Saudi-Houthi theatre and Iraq-origin attacks on Saudi infrastructure are linked by reported alignment and spillover, but the packet does not establish a single operational command. The same distinction applies to Russia's war against Ukraine and Russian-linked hybrid activity in Europe. Broad regional patterns are useful for forecasting exposure, but individual attribution requires separate evidence.

Stakeholder implications

US Central Command and the US Navy face a force-protection problem centred on unidentified maritime attacks under restricted escort windows. Saudi Arabia and Saudi Aramco face uncertainty over East-West Pipeline availability and the split between Gulf, Yanbu and Red Sea export routes. Commercial carriers and insurers face a combined Hormuz and Bab al-Mandab risk premium, particularly for Saudi-linked vessels.

Poland and NATO face a higher warning burden around western Ukrainian strikes, but no confirmed NATO territorial incident occurred on 13 September. Ukraine's rail and port operators remain exposed to further attacks near European supply routes. Taiwan's energy planners face a different warning problem: Chinese patrols and vessel questioning indicate coercive potential, but not an active blockade. UN agencies in Yemen and Sudan face further displacement and access pressure. UN Security Council sanctions decision-makers face an unresolved nationwide Sudan embargo question before 9 October.

Indicators & Warnings

1. Repeated Hormuz attacks: UKMTO reports two or more additional commercial-vessel strikes, fires, crew evacuations or abandonments within seven days. This would confirm a sustained maritime attack pattern. Collection frequency: twice daily. Lead time: hours to days. A signed route protocol and no new incident through 28 September would weaken the assessment.

2. Severe Hormuz constraint: Kpler or Lloyd's List records commodity-vessel crossings below 25 per cent of the pre-conflict baseline for three consecutive days. This would confirm continued severe disruption. Collection frequency: daily. Lead time: three days. Three consecutive days above 50 per cent, with no new UKMTO incident, would weaken the baseline.

3. Operational Iran-Oman arrangement: Oman, Iran, the International Maritime Organization or Gulf port authorities publish a signed protocol specifying routes, fees, escorts and dispute procedures, followed by a 30 per cent rise in AIS-visible tanker movements over three days. This would indicate operational deconfliction rather than diplomatic signalling. Collection frequency: daily. Lead time: immediate after publication.

4. Houthi consolidation: Commercial satellite imagery shows armed personnel, fixed positions or logistics activity at Mokha, Dhubab or Mayyun on two acquisitions at least 48 hours apart. This would support partial tactical control. Collection frequency: every 48 hours. Lead time: two to seven days. A verified Yemeni government recapture would weaken the assessment.

5. Pipeline causation: The Saudi Ministry of Energy, Saudi Aramco or Iraqi authorities confirm physical damage, identify repairs or publish a restart schedule. A delayed restart after visible repair activity would support attack-related damage. Immediate restart without repairs would support a precautionary shutdown. Collection frequency: daily. Lead time: days to two weeks.

6. Western Ukraine strike continuation: Ukraine's Air Force, Ukrzaliznytsia or regional authorities verify two or more attacks on rail, port, fuel or logistics sites within 10 kilometres of Polish border crossings by 28 September. This would strengthen the deliberate-western-logistics assessment. Collection frequency: after every major air-raid period. Lead time: hours to days.

7. NATO spillover: Poland's Defence Ministry, Romania's Defence Ministry or NATO confirms Russian debris, airspace violation or casualties inside Poland or Romania. This would have direct NATO implications. A Moldovan airspace violation would indicate regional spillover but would not itself create NATO territorial considerations. Collection frequency: continuous. Lead time: minutes to hours.

8. Sudan deterioration: UN OCHA, WHO or IOM reports two or more new drone strikes on hospitals, markets, water systems or aid warehouses, or more than 25,000 new Blue Nile displacements in seven days. This would confirm worsening humanitarian conditions. Collection frequency: weekly. Lead time: one to seven days.

9. Nigerian military cohesion: Niger's CNSP or Defence Ministry reports two or more senior command dismissals, officer arrests, armed incidents at Niamey airport or military bases, or a new Russia-linked security deployment. This would confirm worsening cohesion. Collection frequency: daily for the first week, then weekly. Lead time: days to weeks.

10. European hybrid attribution: German, Polish, Slovak, Estonian or Bulgarian authorities publish forensic, financial, device or court evidence linking a named suspect to a Russian intelligence service. This would strengthen attribution beyond the broad activity pattern. Collection frequency: weekly. Lead time: weeks.

Alternatives

• Baseline, severe Hormuz constraint without complete closure, likely at 60-80 per cent: UKMTO's incident report, two protected US transit windows and continued tanker avoidance support this outcome. Continued vessel movement and inconsistent traffic counts argue against complete closure.

• Physical or enforced Hormuz closure, unlikely at 20-40 per cent: Repeated strikes and Iranian blockade language provide a path to closure, but no source establishes that all passage has stopped. UKMTO reporting no commercial passage for 72 hours, verified obstruction or closure notices from multiple Gulf maritime authorities would change the assessment.

• Unidentified or accidental maritime attack rather than a deliberate US-Iran exchange, roughly even chance at 40-60 per cent for the individual incidents: UKMTO confirms the tanker strike but not the attacker, and the Qeshm account is internally inconsistent. Forensic debris or a credible investigation would discriminate between this alternative and deliberate state or proxy action.

• Baseline, Ansar Allah presence and partial tactical control, likely at 60-80 per cent: Multiple major-media and Yemeni accounts support movement into or control of Mokha, Dhubab and Mayyun, but not uncontested administrative control or full Bab al-Mandab closure.

• Ansar Allah presence without consolidated control, roughly even chance at 40-60 per cent: Government counter-attacks, conflicting withdrawal accounts and absent persistent imagery leave open the possibility that Houthi forces hold exposed positions without stable control. Repeated fighting at the sites and continued government administration would support this alternative.

• Ansar Allah secures the full Yemeni Red Sea coast and effectively closes Bab al-Mandab, unlikely at 20-40 per cent: The reported advance provides greater maritime leverage, but there is no verified map showing full coastal control or all-vessel interdiction. Three days of near-zero commercial crossings and verified control of both sides of the strait would support this alternative.

• Baseline, Russian strikes target western Ukrainian logistics rather than Poland, likely at 60-80 per cent: Russia's stated target set, impacts on rail and port infrastructure and Poland's report of no airspace violation support this outcome.

• The passenger train was the specific intended target, unlikely at 20-40 per cent: Ukrainian railway reporting raised the possibility of a diplomatic-train target, but the packet provides no corroborating target-selection evidence. Further strikes specifically against passenger services or weapon telemetry naming the train would strengthen this alternative.

• Accidental Russian spillover into Poland or Romania, unlikely at 20-40 per cent: The impact sites were close to Poland and previous drone incidents affected Romanian and Moldovan airspace. No Polish or Romanian violation, debris or casualty was reported on 13 September. The prior briefing quantified deliberate Russian attack on NATO territory but not accidental spillover, so the current change is qualitative rather than a measured probability increase.

• Pipeline shutdown primarily precautionary rather than damage-driven, unlikely at 20-40 per cent: Saudi statements describe precautionary action, while imagery and other reporting indicate fire or damage near pumping stations. Aramco confirmation of physical damage and a delayed repair-based restart would support attack causation; immediate restart without repairs would support this alternative.

Outlook

From 14 through 28 September 2026, US-Iran maritime confrontation is very likely to continue, with moderate confidence. Hormuz traffic is very likely to remain far below the pre-conflict baseline, while complete closure is unlikely. A limited maritime arrangement is roughly an even chance, but a durable regional settlement is unlikely.

Ansar Allah is likely to retain a presence and partial tactical control at Mokha, Dhubab and Mayyun through 28 September. Bab al-Mandab access is likely to remain selectively constrained, particularly for Saudi-linked vessels. Further Saudi-Houthi exchanges and pressure on Saudi energy infrastructure are very likely by 14 October.

Russia and Ukraine are very likely to sustain high-tempo strikes through 28 September. A decisive Russian breakthrough and a durable ceasefire by 14 October are both unlikely. Russian-linked hybrid activity in Europe is likely to persist through 14 December. An open Russia-NATO kinetic clash remains very unlikely, while an accidental Polish or Romanian incident remains unlikely and more consequential.

China is very likely to sustain coast guard pressure east of Taiwan through 14 December and likely to repeat coercive encounters with the Philippines. An active blockade of Taiwan by 28 September remains very unlikely. SAF and RSF drone warfare, humanitarian deterioration in Sudan, Israeli operations in southern Lebanon and Israeli strikes in Gaza are all very likely to continue. Niger's junta is very likely to retain Niamey through 14 December, while further military tension inside the Nigerien armed forces is likely.