What it comes down to...
... is that older people have knowledge and experience counting against them.
Tech veterans have many years more experience and this can make them expensive. Inexperienced newcomers look like they work twice as hard - and this often obscures that they often achieve half as much.
People who have been working in the tech industry for 20+ years are less likely to entertain nonsense just because it is trendy, buzzword compliant and will impress investors, in large part because they know what the technology does and have been using it since decades before the newcomers are aware it even existed. Unfortunately, the primary objective of business is to make money, not produce a good product. The latter is more a coincidental means to an end.
Put that together, and an experienced worker will appear to not be as productive due to not being an extra keen yes-person, while calling out the obvious insanity of decisions that will not end up being obvious to the inexperienced peers and management for months or years. In the world where the longest term plans are made to the quarterly finance reporting dates, the short termist myopia makes them look like bad hires.
The modern tech industry is full of contradictions. Security is suddenly a buzzword, but only because business leaders have finally taken notice. Once upon a time, sysadmins used to care about it, but the brave new containerised world of curl | sudo bash agile devops utopia has pushed things hilariously past breaking points.
AI is a resurgent buzzword despite the fact that no significant scientific discoveries in the field of machine learning have been made in 20 years, and as Dominic Connor astutely put it, all we have is massively parallel stupidity masquerading as artificial intelligence.
Unfortunately, hiring managers don't want people who can see that something won't work. They want somebody who knows just enough to work hard on solving a problem that is doomed to fail, and keep the investor funding flowing as long as possible.