I believe government transparency is vital to a functioning society. It is what makes governments accountable for their actions, including mass surveillance, military technology, and other activities conducted in the public's name. If taxpayers fund the development of software, they should generally have the right to inspect its source code. Open-source code allows the public, researchers, and independent experts to examine how government-funded systems work, identify potential abuses, and hold governments accountable. There is, however, a legitimate concern that releasing sensitive code could create risks for national security. The development of the atomic bomb demonstrates that technological competition can create dangers even when governments attempt to maintain secrecy. When the Soviet Union developed its own atomic bomb in 1949, the United States lost its nuclear monopoly, intensifying the fear of mutual destruction. Nevertheless, government-funded code does not necessarily mean government-owned or government-controlled code. A company such as Palantir, for example, can receive government funding without its entire codebase becoming government code. The important question is whether software developed with public money should be open-source, and whether any exceptions are justified by genuine security concerns rather than a desire to avoid scrutiny.

That example is a true and legitimate concern, but it assumes that making code public automatically makes the system less secure. Open-source software can actually be more secure because independent cybersecurity researchers can inspect the code, identify vulnerabilities, and report them. Keeping code secret does not prevent attackers from finding weaknesses. In fact, attackers may discover vulnerabilities without the government knowing about them. The better approach is therefore to make government-funded code open-source by default, while allowing narrowly defined exceptions where publishing the code would create a demonstrable risk to public safety or national security. To put it in better words: If the government claims that code cannot be released because of security risks, it should have to explain and justify that exception. Otherwise, ‘security’ becomes a convenient excuse for governments to avoid scrutiny.

Well, I don't think the compromise has to be either total secrecy or publishing everything immediately. You could release older versions of government-funded code once they're no longer in use, along with information about vulnerabilities that have already been fixed. That would still give the public and researchers transparency without exposing the current system to the same risks. I'm not saying we should dump every piece of government code online at once like I'm Edward Snowden. I'm saying there should be a way to balance transparency with legitimate security concerns.

I understand that concern, especially with defense systems. But I think the important question is whether keeping the code secret actually makes it more secure. If only a few approved people can inspect it, vulnerabilities can potentially go unnoticed for years. Open-source code gives independent security researchers more opportunities to find those weaknesses. That said, I wouldn't argue that every defense system should immediately be published. If releasing the code would genuinely expose sensitive capabilities or create a serious security risk, there should be an exception. My argument is that secrecy should have to be justified, rather than automatically assumed just because the government funded the software.