I used to work for a large multinational group whose core financial systems were written in COBOL during the seventies. I was told need to think of making the core systems future proof was identified as the company grew through the eighties and nineties, but never addressed. As time went by this attitude turned from one of "let's try and get on top of this" to "it can't be done because it's too risky / complex / old"...

The net result of this is that this ignorance spilled into the attitude towards technology across the group. I found it very demoralising as you always had to fight to convince non-technical staff why they needed to address technical debt that was 10/20/30 years old. In the end I managed to undertake some big and interesting projects successfully in spite of the "you'll never achieve that without a disaster" attitude, but the core COBOL system were just ignored as if somehow it'd go away (the reality was much more political and complicated than that, as it is with all multinationals...)

In the end I left for a position where people wanted me to build or rebuild things, rather than just spring the leaks around the problems. My point here is that unless you wield obvious experience or weight in this arena (as the "COBOL-legends" undoubtedly do) you'll have a job convincing people that it needs doing, at least until it goes up in flames.