After working in a couple unicorn FinTechs:

They're just like traditional companies, but with much less oversight, attention to regulation and transparency.

At the beginning there is plenty of support channels, but that's because of marketing and because there's investor money. But as soon as the money gets tight, people start suggesting dark patterns and everything becomes "you must contact us to do X".

Not only that but there is way less auditing, as technical auditors that deal with fintechs aren't really ready to deal with anything made after 1990. That's you, Deloitte.

Also regarding security, what I saw in practice was that everyone has access to absolutely everything and could do anything. Everyone but the lowest support people can transfer money from anywhere to everywhere, change passwords, view and edit personal information, collect private data. Customer personal data is sent around in Excel files in email like it's candy. There was SO MUCH logging that seeing suspicious employee activity was basically impossible without having a complaint from the customer itself and a thorough investigation (which rarely happened).

Also: AI and Data Science are mostly people running one or two queries per week in the production DB, exporting to CSV and calling it a day.

And I don't want to dox myself but: a popular German FinTech with "AI" in its name has way less automation than a 5 person startup. Every single operation is triggered manually and is super error prone. And such operations involve 20, 30 records, when there are 10 million in the database.

Recently there was allegedly a kerfuffle with a german Fintech bank banning hundreds of users because of suspicion that they had gang relations. Well guess why.