"UX in big tech = user exploitation" - this is a quote from a recent article by Mark Hurst, and I thought it really nailed it.
Amazon, FB, Medium etc still use dark patterns to trick us into giving away personal details, or stop us from cancelling subscriptions specifically through deceiving UI copy. For example, they often use euphemisms to disguise negative things:

Facebook intentionally package their intrusive, data-hungry activity as a feature that benefits users. Cross-website tracking and hidden pixels becomes your ‘Ad preferences’. There’s no clear mention of tracking or mining – it’s all euphemisms and positive spin that masks what’s happening in order to make users feel in control
Amazon’s cancellation page is an example of using positive wording to mislead. Amazon use ‘Prime Benefits’, or ‘Benefits’ as a veil for cancellations. So instead of a negative ‘Cancel Membership’ page, you get the more positive ‘End Benefits’ page.

Cancel? That’s not a word 🤑
Blogging platform Medium often persuades writers to publish stories behind a paywall. However, due negative associations with the word ‘paywall’, they often replace it with more positive terms and phrases such as ‘Partner Program’ or ‘Meter my story’

A more indie-related example is when Baremetrics was sold by Josh Pigford – straight away, the acquirers added popups to prevent customers from cancelling:
Now that money is prioritised over the product, you have to have a call with ‘non-salesperson’ Brian before cancelling your subscription.
As indie makers, pressure can build up to make money, and it can push us towards the use of dark patterns. However, at what point do they erode trust and ruin the product?
If you're interested in the topic, I wrote a giant case study that goes into a lot more detail, with many examples:
https://prototypr.io/post/we-value-your-privacy-at-about-0-50-dark-patterns-in-ui-copy/