A defence system is bought to be operated, corrected and adapted long after it has been delivered, and where the system depends on software, much of that work is done in code. The armed force that operates the system rarely writes that software itself. It depends on the supplier, or on a subcontractor further down the chain, to fix faults, adapt the code to new hardware or threats, and keep the system usable. That dependence is tolerable as long as the supplier exists and performs. It becomes a question of law the moment the supplier enters insolvency proceedings. At that point two bodies of rules meet: the insolvency law of the state where the supplier is established, which decides what happens to its unfinished contracts and its assets, and the copyright and contract law that decides who may use, copy and change the software. A European ministry of defence that has paid for a system may discover that it owns the hardware but not the right to change the code, that the source code was never delivered, or that the build environment needed to turn the code into a working program sits on the supplier’s premises under the control of an insolvency administrator. The practical question for European buyers is therefore not abstract. When the supplier of the software in a European defence system fails, what decides whether the armed force can go on maintaining and modifying it: the insolvency law of the supplier’s state, or the rights and materials it secured in the contract before the insolvency?