The Forecasting MachineOpen interactive report

Will Iran conduct a deadly attack on a military base in a NATO country before the end of 2026?

This question forecasts whether Iran will carry out a lethal attack against a military installation located within a NATO member state. Recent intelligence reports indicate a shift in Iran's strategy towards targeting 'harder targets,' including military sites, and an increase in hybrid and sabotage attacks across Europe. NATO has intercepted Iranian missile launches towards Turkey, and while NATO as an organization has avoided direct involvement in the US-Iran conflict, individual member states have been used for logistical support, and some bases have been targeted by Iranian drones or faced security incidents. The timeframe for this forecast is before the end of 2026.

Forecasted on

Forecast probabilities

  • Yes: 3.4%
  • No: 96.6%

Reasoning

Iran has a 3.4% chance of carrying out a deadly attack on a military base in a NATO country before the end of 2026; No is 96.6%. The danger is real, as the war remains active and Iran-linked plots have reached Europe, but the remaining window is short and the evidence so far points to a pattern of failed or nonlethal attacks on NATO territory.

Yes — 3.4%: Escalation could breach the boundary

The conflict has not settled into a durable ceasefire. Iran and the United States have exchanged strikes again, including Iranian drone attacks on U.S. facilities in Kuwait and Bahrain and a U.S. strike on Qeshm Island [e874, e882, e888]. Iran has also threatened bases that support attacks on its territory, and reports have identified facilities in southeastern Europe, including Bezmer and Akrotiri, as potential targets [a17, a19].

The threat is not only military. A suspected Iran-directed plot against a UK airbase failed, police arrested suspects near RAF Fairford, and the alleged organizer has faced charges in the United States [e876, e878, e879]. Suspected drone activity over RAF Lakenheath, cyberattacks attributed to Iran, and disrupted plots in Greece and North Macedonia underline that Iran-linked activity can reach European targets [e885, e872, e886]. If a new U.S.-Israeli campaign begins, retaliation could expand; Israeli leaders have pressed for one, and U.S. officials have assessed possible Iranian responses [e893, e894].

But a deadly attack requires more than hostile intent: it must reach a NATO-country military base and cause a death. The reported UK plot failed, the Lakenheath incident caused no harm, and Iranian missiles aimed at Turkey were intercepted [e879, e885] [a9, a10]. Attacks on bases in Bahrain or Kuwait do not meet the question’s geographic condition. Even the Akrotiri drone incident caused no deaths and was attributed to Hezbollah proxies [a17, a19].

No — 96.6%: A short window and a clear pattern of restraint

The most informative evidence is Iran’s conduct inside Europe during the wider conflict. Its reported hybrid attacks have focused on arson and explosives against Jewish institutions and US-linked targets, with intelligence describing the campaign as designed to avoid casualties [a7, a8, a18]. This does not prove Iran would never accept fatalities, but it suggests Tehran has reason to distinguish intimidation and disruption from a lethal attack that could provoke a much larger response.

The military option also faces obstacles. Iranian missiles capable of reaching southeastern Europe reportedly have accuracy and payload limitations at those distances, and inventories have been depleted. NATO defenses have intercepted Iranian missiles aimed at Turkey, while the attempted long-range strike on Diego Garcia failed to hit its target [a17, a19, a23]. Host-nation security services have also shown they can disrupt plots before they reach a base [e876, e886, e927].

Finally, the calendar matters. Only about six weeks remain in the forecast period. President Trump has said he will wait until after January 1 before deciding on further military action, though he has not ruled out acting earlier; meanwhile, mediators continue to carry messages between Washington and Tehran and regional governments are trying to prevent renewed full-scale war [e877, e902, e931, e906]. That leaves room for an unexpected escalation, but makes a major new trigger before year-end less likely.

Why the balance favors No

The case for Yes rests on a genuine escalation risk: an active war, plausible European targets, and evidence of Iranian-linked operations near military facilities. The stronger evidence for No is that Iran has not yet caused a fatality at a military base in a NATO country, its European activity has so far been thwarted or calibrated to avoid deaths, and the remaining time is brief. Those forces justify a low, but not negligible, 3.4% probability of Yes and 96.6% for No.

Update summary

Trump said he would wait until after January 1 to decide on further military action, though he left open an earlier decision [e877, e902]. Messages between Washington and Tehran continue through mediators, but talks remain stalled over Iran’s uranium stockpile [e931, e933].

Cited evidence