Before Article 5, NATO Must Be Ready to Defend Poland

“Accidental” is the explanation Polish Prime Minister Donald Tusk warned NATO to anticipate. On September 17, 2026, he told parliament that Russia was planning drone or rocket strikes against countries supporting Ukraine, including Poland, while preparing to portray them as mistakes. That is an intelligence-based warning from Poland’s government, not public proof that a particular attack has been ordered. It nevertheless poses an urgent problem: how should allies protect Poland while determining what happened and whether a wider military response is justified? (apnews.com)

NATO has already confronted parts of that problem. On August 12, allies discussed fresh airspace violations in Poland and Romania, blamed Russia and reaffirmed their support for Ukraine. Their statement emphasized existing defensive action rather than announcing an Article 5 response. The space between an incursion and collective-defense declarations is therefore where much of the alliance’s practical work already happens. (nato.int²)

I would distinguish three decisions that require different evidence: stopping an immediate threat, positioning forces against the next one, and using force against those responsible. Treating them as a single escalation decision invites either paralysis or overreaction. NATO’s published air-defense policy already provides a useful foundation: senior commanders allocate resources, while tactical units execute within delegated authority. The task is to make that arrangement work quickly without turning every destructive incident into a predetermined reinforcement or retaliation order. (nato.int³)

The treaty leaves room for this distinction. Article 4 allows any ally to call consultations when it considers its territory, independence or security threatened. A Polish request would bring the issue before the North Atlantic Council, the alliance’s principal political body. NATO makes collective decisions by consensus, but Article 4 is neither a required preliminary step toward Article 5 nor a promise that consultations will lead there. Leaders can consult while authorized defensive operations continue. (nato.int⁴)

Article 5 establishes a different obligation: an armed attack against one ally is treated as an attack against all, and each must assist through action it considers necessary, potentially including armed force. It does not order every member to launch the same military operation. NATO’s current explanation says the obligation arises when an ally suffers an armed attack and requests or accepts collective action; the treaty prescribes no single activation procedure. A Council declaration coordinates and communicates the response, rather than serving as the first permission to protect Polish territory. (nato.int⁴)

In the immediate encounter, NATO’s operational framework centers on detection, tracking, identification and predicted flight paths. Those assessments support interception and early warnings to national authorities. They need not settle why a weapon crossed the border. During the September 10, 2025 drone incursion into Poland, previously delegated authorities allowed aircrews to engage threats, NATO’s supreme commander in Europe, General Alexus Grynkewich, subsequently explained. Political control had been exercised in advance, allowing action on the shorter timeline available to defenders. (nato.int³)

Responsibility requires a separate investigation. The fatal explosion in eastern Poland on November 15, 2022 illustrates why consequences cannot identify the attacker. The next day, NATO said its preliminary assessment pointed to a Ukrainian air-defense missile launched against Russian cruise missiles. It reported no indication of a deliberate attack on Poland, while the investigation continued. NATO also held Russia ultimately responsible for the circumstances created by its war. Those judgments could coexist: responsibility for starting the war did not establish that Russian forces had fired the weapon that struck Poland. (nato.int⁸)

The 2025 incursion supplies a different lesson. Poland received Article 4 consultations on September 10; NATO announced Eastern Sentry, an effort to strengthen eastern-flank defenses, on September 12. Contributions came from France, Germany, Denmark and Britain. Grynkewich said he had issued the start order under existing authorities, although integrating the contributions would take longer. Russian intent was still under assessment. I read that sequence as evidence that reinforcement and unresolved intent can coexist, not evidence that every future incident should produce the same deployment. (nato.int)

There is an appealing argument for making additional protection the default after repeated or destructive hostile incidents: an attacker should not gain time by claiming an accident. I share the concern, but repetition does not tell commanders which system to move or what becomes less protected when it moves. Poland’s Defense Ministry made the allocation problem explicit on August 19, 2026. Its defenses do not consist of permanently stationed missile batteries around every major city; assets are assigned according to threats and priorities. It withheld detailed deployments and readiness information for operational reasons. A credible response must protect against the assessed next attack, rather than merely look larger after the last one. (gov.pl¹⁰)

Public evidence establishes real capability, with important limits. Poland conducted Patriot live firing in September 2025. NATO’s Ramstein Flag exercise in June 2026 practiced integrated air and missile defense and rapid information-sharing with more than 200 aircraft and aerial assets. These are meaningful demonstrations of equipment and cooperation. Neither account establishes how many ready interceptors Poland has today, how long stocks would last under repeated attacks, or how quickly every reinforcement could become operational. I would not turn an exercise report into a claim of comprehensive coverage. (gov.pl¹¹)

Sustainment is also receiving attention. Britain announced the opening of a missile repair and logistics facility in Poland on September 7, 2026, intended to speed maintenance, support and training for Polish air defenses. That is relevant to keeping systems usable. It still supplies no public tally of ready ammunition. Before ordering a wider deployment, leaders should demand an operational comparison of protective benefit, ammunition, integration time and coverage sacrificed elsewhere. Both reinforcing and declining reinforcement deserve recorded reasons. (gov.uk¹³)

For an apparently accidental, isolated strike, I would favor immediate protection against any continuing danger, investigation, targeted defensive adjustments and diplomacy. Fatalities would increase the urgency of rescue and accountability without independently establishing an armed attack by Russia. The 2022 preliminary finding shows why this distinction matters even when the human consequences are grave. (nato.int⁸)

For an ambiguous or recurring incident, I would favor rapid consultations and reversible precautions: additional surveillance, adjusted patrols, raised readiness and selective reinforcement. NATO’s policy includes rapid repositioning, early warning and physical protection alongside interception. Uncertainty about intent should not block these measures, but deployment choices should remain tied to the expected threat. Repetition should change the assessment of future danger without replacing it. (nato.int³)

For a clearly attributable armed attack, I would support Article 5 assistance tailored to Poland’s needs. The response could combine military protection with intelligence, logistics and other support; allies need not contribute identically. Any proposed strike against the attacker should receive a separate assessment of responsibility, necessity, proportionality and escalation risk. Article 5 is grounded in self-defense, not a standing license for punishment. (nato.int⁵)

Tusk’s warning deserves preparation that does not depend on accepting either Moscow’s explanation or Warsaw’s initial interpretation of a future incident. I would want NATO able to stop the next threatening weapon while investigators and leaders establish responsibility for the last one. An adversary’s claim of accident should buy it neither a pause in protection nor control over the alliance’s decision to widen the conflict.