Can AI End the Ukraine/Russia War?

“That sounds incredible!” is one of the most common responses I get when I describe “The End of Conflict” to someone.¹

But, of course, “incredible” can mean “difficult to believe” and I can see their brain processing, looking for a test case. A question about Ukraine/Russia often follows.

And, no, artificial intelligence is not yet some magical machine that can instantly turn war into peace and Vladimir Putin into Mother Teresa. But it is becoming one of our most powerful tools in conflict resolution, helping build lasting solutions even in entrenched situations.

Let’s see some of the things it can do in the Ukraine/Russia war.

When is a war ripe for peace?

For peace talks to have any chance of success, the timing is critical. Ripeness theory was developed by William Zartman as a way of understanding that the right circumstances need to be in place if the negotiation is to be successful.²

But Project Didi has developed a machine learning model for evaluating how ripe a conflict is for negotiation.³ It can then recommend what still needs to be in place, how to put them in place, and then how to intervene once they are in place.

This is a major leap forward for peacetech – the rapidly growing field using AI and other new technology for peacebuilding and conflict resolution.

Data-driven peace

War is obviously a highly complex, often chaotic, phenomenon, making it very difficult to study in any systematic way. But there are now databases that can help, like the University of Edinburgh PeaceRep which contains over 2000 peace agreements from the last 30 years and a whole host of AI-supported tools to help analyse them.⁴

The Washington think-tank CSIS (Centre for Strategic & International Studies) used one of the most well-known databases, UCDP, to analyse 333 peace agreements from a similar period and then related the findings to the Ukraine situation.^{5, 6}

Their findings? Successful peace deals tend to include:

Involvement of local communities and key stakeholders
Land-for-security compromises
Third-party security guarantees
Effective monitoring and verification
Strong plans for economic reconstruction and recovery.

Negotiating trade-offs and compromise

So, from this, we can see where a peace agreement might lie. Both sides will need to compromise but they may think it’s worthwhile.

War fatigue means that a majority of Ukrainians are now willing to concede land for peace and a similar majority in Russia also favour talks.^{7, 8}
European and/or US troops could still provide security whilst meeting a Russian demand to change the 2008 promise to allow Ukrainian NATO membership in the future to disallowing it unless attacked.
And prosecution of war crimes would probably need to be dropped but the $300 billion of Russian central bank reserves mostly held in the Euroclear Bank in Belgium could be used to pay for much of the country’s reconstruction.

The AI model was also able to generate a number of alternative peace agreements, possible trade-offs and identify what might cause a breakdown – like domestic factions sabotaging any agreement and the extreme lack of trust between the two countries.

Think you’re a good mediator? They’ve got a simulator which you can play with. It’s a lite version of the real thing but it can give you a sense of what it can do.

Building a sustainable peace

But AI can help in implementation, too. It's used to manage drones and analyse satellite imagery to monitor the agreement; and social media analysis, important in ripeness evaluation, can also be used to predict potential violations.^{9, 10}

And AI-supported deliberative dialogue platforms, like Remesh and Polis, optimised for agreement, have been used successfully to build bridges even in such conflict zones as Libya and Yemen. They can be used to develop peace agreements that all sides can support.

What’s more, true peace means more than a ceasefire. If we want it to last, it needs to be more than a begrudging tolerance and, ultimately, move to a healthy relationship based around trade, shared interests and positive interactions. Peacetech can help here, too, supporting cross-border cooperation and community-led reconciliation.¹¹

Conclusion

Let’s be clear, the Ukraine/Russia war is a tough one to resolve and it’s unlikely the two countries are going to be best friends this year.

But don’t judge AI’s future impact on today’s capabilities. All of the above is already happening now and the tools, already useful, will only get better. Expect peace in the region at some point and AI and peacetech are going to play a big part in making that happen sooner rather than later.