What Could a Trump Peace Plan in Ukraine Look Like?

In the aftermath of the Gaza deal, the Trump team is renewing its push to end the war in Ukraine. In the last week, it has emerged that the president has been negotiating behind the scenes and is set to meet with Russian President Vladmir Putin in Budapest in the near future. This deal comes on the back of renewed pressure against Russia both in terms of U.S. military aid to Ukraine—potentially including long-range missiles like Tomahawks—and financially through pressure on Russian oil sales to India. At the same time, the European Union and the United Kingdom see a path forward for using frozen Russian assets to finance Ukrainian weapons purchases, and the United States is talking up “victory tariffs” to compel Moscow to end hostilities. Taken together, Russia is facing the prospect of increased Western backing of Ukraine at the same time that its economy—both due to sanctions and Ukrainian attacks on oil and gas infrastructure—appears on the verge of collapse. Yet, what could the plan look like?

This installment of Critical Questions reviews findings from the last year published in the CSIS Futures Lab Strategic Headwinds series on the opportunities and challenges of negotiating—and implementing—a peace deal in Ukraine. This series combined expert surveys and responses generated by an AI model trained on historic peace negotiations. Both experts and AI agents converge on the view that avoiding de jure recognition of Russian-occupied territories will be critical for Ukraine, and that Ukraine’s NATO membership prospects would need to be taken off the table for the foreseeable future.

Q1: What’s the most likely deal according to experts?

A1: According to the experts surveyed for CSIS’s Strategic Headwinds series, the most likely settlement looks less like a grand bargain than a managed freeze of the current conflict—one that de-escalates violence without conceding principles that would turn modern land grabs into precedent. In practice, that could mean shelving formal NATO membership for now, narrowly calibrating sanctions relief to verifiable Russian behavior, and tightly limiting any use of frozen Russian assets to reconstruction under international oversight. It would stop short of de jure recognition of Moscow’s seizures while acknowledging the de facto line of control as a temporary military reality—not a political settlement. Paired with a surge to rebuild Ukraine’s defense-industrial base and explicit noninterference in its domestic politics, this path seeks to close Moscow’s windows for coercion while avoiding concessions that reward aggression. The sticking points are predictable: credible guarantors and any peacekeeping presence, both of which the Kremlin will resist unless the terms favor its freedom of action.

What is less likely is a maximalist approach that forces capitulation by either party. Full Russian withdrawal with comprehensive accountability for war crimes is morally compelling but strategically improbable in the near term; so too is a Ukrainian renunciation of sovereignty or recognition of annexations that would legitimize territorial revisionism. Formal NATO accession or mutual recognition of altered borders similarly cross red lines that neither side is prepared to accept. In short, the workable space is narrow: reduce the rate of killing, harden Ukraine’s long-term deterrent, and avoid “peace” formulas that convert aggression into a new normal while setting the stage for post-conflict reconstruction.

Q2: What’s the most likely deal according to AI models?

A2: AI models are not entirely aligned with the human experts. To assess how AI models approached the peace process, the CSIS Futures Lab used GPT-4o-mini, prompting the model to take on the roles of national leaders and answer survey questions based on the information available, which largely consisted of news articles from multiple countries on the peace process. Each question was asked 1,000 times, and the results were averaged to produce the final scores.

AI-derived assessments cast Moscow as far less flexible than most human experts on Ukraine’s rearmament and defense-industrial revival. Where practitioners see room for calibrated Western–Ukrainian security cooperation, the models treat it as a near nonstarter. They also flag Western commercial access to Ukrainian resources as close to a red line. This last point stands in direct contrast to ongoing efforts to link critical minerals to lasting U.S. involvement in Ukraine and any peace deal. On the Ukrainian side, the models harden around territory. Any move that legitimizes Russian control—formal recognition or continued occupation in places like Zaporizhzhia—reads as unacceptable. The models are also tougher on frozen Russian assets, treating release as off limits rather than a chip to trade.

There is still meaningful convergence between the AI models and human experts. NATO membership is off the table for now. Sweeping war-crimes accountability is unlikely in the near term. And while the models and experts expect no direct use of frozen Russian assets to fund reconstruction, they point to a workable middle ground. Those assets could back loan guarantees or similar instruments that channel capital into rebuilding while keeping pressure on the Kremlin. That approach preserves leverage and creates an accountability mechanism without normalizing aggression. Both communities also align on two hard guardrails: no formal limits on Ukraine’s defense industry and no de jure recognition of Russia’s occupation.

Q3: What’s the best chance for a deal?

A3: The best chance for a deal would likely center on removing non-negotiables from the table and freezing the conflict along current frontlines, without granting de jure recognition of Russian occupation. While NATO membership would remain out of reach for Ukraine, Russia would likely have to accept some forms of security cooperation between Ukraine and Western partners. Russian assets are unlikely to be used directly for Ukraine’s reconstruction; however, this issue could become a negotiable element within a broader agreement, potentially traded for partial sanction relief or other economic concessions.

The most difficult phase would begin after a frozen conflict takes hold—ensuring that fighting does not resume. This will hinge on the question of international guarantors and potential monitoring forces, which Russia is expected to oppose.

In addition, economic statecraft—which is increasingly playing a driving role in a new Trump doctrine—will be at the center of any deal. Trump will be negotiating not just about the war, but also about how Ukraine fits into his vision of a global economic order that is defined by critical minerals and infrastructure abroad that supports American prosperity. That means Ukrainian geography, not just Russian battle positions, will be part of the calculus, including how to use Ukraine’s network of riverine ports, Black Sea ports, rail lines, and cheap energy as opportunities for American capital to drive reconstruction.

Q4: What’s the current risk of failure?

A4: Any peace deal will involve a ceasefire, and most ceasefires are fragile. Based on historical analysis, 31 percent of interstate wars end in a stalemate under ceasefire agreements, which halts large-scale violence but leaves underlying disputes unresolved. Worst still, most ceasefires experience minor failures within 10 days, with larger failures tending to occur between 65 and 193 days. The involvement of external monitors tends to reduce large-scale violence but not minor clashes. These observations mean that any ceasefire in Ukraine is almost certain to fail absent some external monitoring and security guarantee.

Past ceasefires in Ukraine did not hold. Kyiv accuses Moscow of violating 25 ceasefire agreements since 2014. The Minsk I Protocol (2014) and subsequent memorandum quickly broken down and paved the way for renewed offensives. Moscow has historically used ceasefires more as a way to position forces on the battlefield than to seek long-term peace. Even the Minsk II agreement only slowed fighting to the contact line in the Donbas while setting the stage for the 2022 invasion. This track record further reinforces the historical trend: Lasting ceasefires need a mix of monitors and security guarantees.

Q5: How else could a successful peace deal fail?

A5: Just because guns stop doesn’t mean politics ends. If anything, most post-conflict settings are prone to political crises and periods of economic boom and bust cycles. Wars also produce sustained economic shocks that can compound domestic political infighting after conflict. This means that any peace deal must come with commitments to economic reconstruction, elections in Ukraine, and sustained EU and U.S. support for rebuilding society.

A potential Trump peace plan would likely seek to freeze the war rather than end it, trading formal recognition and accountability for immediate stability and economic opportunity. Yet without credible guarantees, strong monitoring, and sustained reconstruction commitments, such a deal risks becoming just another temporary pause in Europe’s most dangerous conflict since World War II.

Benjamin Jensen is director of the Futures Lab and a senior fellow for the Defense and Security Department at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Yasir Atalan is a data fellow in the Futures Lab at CSIS.