The Battle Does Not Stay On the Field: The Limits of Dual use

What we learned 

Both researchers described the Palestinian territories as a "living lab" (Ibrahim's term) where AI surveillance technologies are developed and deployed on a civilian population with no meaningful ability to refuse. In the West Bank, this includes facial recognition at checkpoints, AI-powered drones, and biometric data systems. In Gaza, drawing on soldier testimonies documented by independent investigative journalists, Ali described AI targeting tools operating at a scale that outpaced meaningful human review. 

The broader argument was that these patterns are not unique to Palestine. Technologies developed in contexts without meaningful legal constraint do not stay there. They are marketed, exported, and adopted: often by the same cloud infrastructure that European institutions use every day. 

On accountability, the researchers made a strong point: responsibility does not disappear when something goes wrong, but it distributes itself across technical, legal and institutional layers until it becomes practically impossible to locate.

On dual use

For me, the challenge to the concept of dual use was also striking. The conventional framing holds that AI technologies are essentially neutral: tools that can serve good or bad ends depending on who deploys them for what. But if a technology is built from the very beginning within a context of military control and collective punishment, that distinction collapses before it can even be applied. The values embedded in the design travel with the product. Responsible use cannot simply be bolted on afterwards.

A personal reflection 

I want to be honest: writing on this guest lecture proved nearly impossible to do in a diplomatic manner. And that difficulty is itself worth reflecting on. 

How do we hold our emotions alongside our academic perspective? Behind every data point in Ibrahim and Ali's presentation is a child, a parent, a family. That is true in Palestine. It is true in Israel. It is true in Ukraine and Russia, in Iran, and in the many other places where AI-driven systems are being deployed in war and conflict right now. Every one of those lives matters equally. And every one of those tragedies deserves to be named as such: not instrumentalised for a political argument, but simply mourned. 

I am deeply grateful to researchers like Ibrahim and Ali, who work in conditions most of us cannot imagine and who choose science as their weapon: not to destroy, but to build a more humane future. That takes courage of a rare kind. And it gives me genuine hope that peace is possible. 

From that place of mourning, the dual use question becomes something much larger than a technical or legal debate. It becomes a question about the world we want to live in. Military technologies do not stay on the battlefield. They enter our cities, our institutions, our public safety systems and, quietly, our sense of what is normal. Once we accept that certain uses of AI are inevitable in conflict and war, we have already begun to shape what we will accept closer to home and in more mundane public safety questions.  

That is the deeper reflection Ibrahim and Ali's work left me with. Not only what is happening in specific places, but what it reveals about the systems we are all building and living with — and the responsibility that places on researchers, institutions and societies to ask, clearly and without evasion: what kind of world are we choosing?

At the close of the session, we from AI MAPS offered to deliver a lecture for students at Birzeit University: a reminder that research is not only about extracting knowledge, but about building relationships across distance and difference.