They don’t need to, OpenAI and Anthropic are running at a huge loss to keep it cheaper to use them than to buy the comparable hardware required to replace them. The only way running open source models is cheaper is if you ignore hardware costs, or if you have a very specific and unusual usage pattern (i.e. constant 100% capacity, and not a clock cycle more or less), or if you go with another cloud provider who are also running at a loss.

hang on a minute, there’s spaces around your em-dash! An AI didn’t write this at all, this is fake AI! I call Shenanigans!

Using established and tested security implementations instead of rolling your own has been considered good practice since forever. It’s not a bad thing.