Very true. And account integrity check pointing is stochastic and more aggressive so at any time there are people being locked out .

One of 20 of your services could lock you out tomorrow and that means you’re blocked from coworkers and family

The root issue/risk is from cascading service dependencies, and I'm 99% sure this is done unintentionally at Apple et al.

Team builds service. Service depends on first party identity/authentication because it's easier.

... Fast forward 20 years, and no one at platform company even understands the dependency graph from a customer perspective anymore. Especially in the case of rare events like account locks.

Consequently, those customers face a sudden Kafkaesque maze of edge cases that don't line up, as the customer service processes people are funneled through are literally incapable of solving the problem.

Which means the entire "normal" customer support apparatus is unavailable to them. (The same apparatus companies aggressive shove all support through)

This is why there should be regulatory requirements for identity platforms mandating the ability to speak to a human who's empowered to fix your issue + an option for customer-choice decision arbitration + continuous random sample audits with penalties for falling below KPIs (timeliness, correctness, etc).

It should literally be illegal for a company to have their banning system 'oops' and then pretend they don't know you.

Because it's only going to get worse as more AI / probably correct methods infuse account security functions.

[flagged]