From the smartphones in our pockets to the cruise missiles on the battlefield, the software that powers our modern world is increasingly built by anonymous volunteers. The open-source movement, which relies on public, collaborative coding to find and fix bugs, has become a massive success. Yet this open philosophy is now raising serious alarm bells as it transforms warfare and disrupts the global artificial intelligence race.
In Ukraine's fight against invasion, cheap consumer drone technology has become a vital asset. However, open-source software originally designed for recreational drones has now been adapted to guide inexpensive Ukrainian cruise missiles. This transition highlights a dramatic shift: advanced military guidance systems that once cost millions to develop can now be downloaded by anyone with an internet connection.
A parallel struggle is playing out in the realm of artificial intelligence. Unlike US tech giants that lock their models behind commercial paywalls and strict safety guardrails, developers in China are championing open-source models. While these open models are slightly behind in raw capability, they are free to download and run locally.
The catch is that open-source AI models are far more difficult to control. The safety features meant to prevent them from generating harmful content are minimal, and what safeguards exist can be easily stripped away. This presents a major headache for regulators who want to audit models for national security risks before release. While the open approach allows developers worldwide to find innovative new uses for AI, it also ensures that as technology makes some of society's problems easier to solve, others become far harder to contain.