"When data is encrypted "end to end" it means that no one - including law enforcement and tech firms - can access the data without the key."

Depends on who does the encrypting, i.e., who controls the software

If the user does the encrypting, then the statement is true. The companies cannot access the data. And the government is not going to ask users to modify their software to get access to data. It would just ask users for their data directly. Users could defend themselves in this situation. A long history of legislation and jurisprudence exists for resolving such requests

But if it's a so-called "tech" company doing the encryption, an intermediary (middleman) forcing users to use the company's software for that purpose, then the government can ask the company to modify the software to allow data access by the government. Thus the quoted statement is false. The companies can access the data, simply by modifying the software

Users cannot defend themselves in this situation. The companies may or may not alert the user when their data is requested, before their data is accessed

The companies "update" their software continually through remote access to users' computers. There is no way for the user to know if the software has been modified to allow data access by the government

Hence the design of so-called "end-to-end encrypted messaging services", using "tech" company controlled software and "tech" company controlled servers, is flawed

And, alas, journalists make misleading statements like the one above

The accuracy of such statements is currently being questioned in litigation against Meta. Stay tuned