Let's assume the worst case scenario: 100% of all current contributers to Bitcoin Core (hereafter Core) has been convinced by some government to try to subvert Bitcoin. This means that all future releases of Core would have some sort of a malicious changes to it.

First, most nodes are running old versions of Core. Subverting future releases of Core could only affect new nodes/users. And new nodes/users could easily run old versions of Core, which woudl happen the instant the subversion were discovered.

Second, the most useful subversions are the most likely to be found. For instance, if a new release of Core would send some of a user's BTC to a government-controlled address, then that would become visible within hours of being released. Users would complain on Github, Reddit, Twitter, Facebook, Stacker News, IRC, Nostr, and many other places.

Third, for any subversions to work, they must be kept hidden. And hiding bugs in an open source project only works if people don't care about the project. Core's development affects people's livelihoods and fortuntes, so a lot of people are scouring it's code for bugs. Moreover, those seeking to steal BTC look for code to exploit. If a backdoor were added to Core, it would be taken advantage of quickly, thus exposing the subversion.

Fourth, if the subversion(s) were discovered, the Core devs would be treated like criminals forever. And others would scour the Core source code to find any possible subversions. It would be somewhat like when the OpenBSD devs started LibreSSL, because OpenSSL had made too many mistakes over the years -- except the OpenSSL devs weren't obviously beig malicious.

Fifth, the changes that the Core devs can actually make to Bitcoin itself are quite small. Soft-forks and hard forks only affect new versions of Bitcoin Core. The rest of the old versions of Core would not follow either.

Sixth, mining pool companies care a great deal about the integrity of Bitcoin. They will not risk losing customers (i.e. the people with mining hardware) by going along with a hard fork created by Core devs.

Finally, consumers (i.e. individuals spending their own money) are the ultimate deciding factor. When a hard fork occurs, consumers will buy one one side of the fork and sell on the other side of the fork. The miners/pools will mine on the side that has the highest price (presumably after a few days). Fortunately, a lot of consumers don't even hear about hard forks until long after one side has been abandoned. For a hard fork to be successful, there must already be a culture of users constantly updating to the latest version (like Monero). Bitcoin users tend to run old versions of Bitcoin Core.

If you need clarification on any of the above, let me know. Corrections are also welcome.