Tutorial 27 · Security and controls: what leaves your computer, and where to

What the video says

1. Security and controls is where you decide what may leave your computer and where to. It lives in the menu, under Management and administration, and every control applies instantly, with no restart.
2. The personal data firewall checks every message before it goes out to an external provider. If it finds personal data, it applies the rule you set for that provider. It ships switched on.
3. The four counters say what it has done: how many turns it handled locally, how many it blocked, how many it pseudonymised, and how many went out as an exception. That last one is what to look at in a review.
4. Below, the rule for each provider. The local ones start out with no filtering, because nothing leaves your computer anyway; even so you can tighten them if that machine is shared with other people.
5. The internet ones start out blocking. That is the opposite of the usual: here relaxing the protection takes a deliberate action from you, it does not come relaxed out of the box.
6. Three rules and no more. Never send it leaves the turn on your local brain. Replace with placeholders sends the message with pseudonyms. And no filtering sends it as it is: that is an exception, and it gets logged.
7. And this button, How I protect your data, opens the full explanation. It is the screen you show a client, an auditor or your compliance officer when they ask what the system actually does.
8. It starts with the path a message of yours follows, step by step. Every time you write, the message passes through a reviewer before it goes anywhere, and that reviewer runs inside your computer, with no connection.
9. Then, what it considers personal data: ten types, with an example of each, how it recognises them, and what they turn into if they go out. Name, ID document, email, phone, postal address, bank account and date of birth.
10. The last three are critical categories, marked never leaves: health, passwords and payment cards. They are not replaced by placeholders and do not go out under any setting, because a placeholder protects who you are, not the fact of being ill.
11. Section three explains how it recognises them, and why you can trust the method: written rules that run on your computer. There is no external service reading your messages to decide whether they carry personal data.
12. When in doubt, it protects. It looks at the whole message, including remembered context and the documents consulted. And it never stores the value it found, only the type: in the logs you see email or document, never the data itself.
13. Section four is the clearest of all: the same message, and what the provider receives under each rule. Blocking, it receives nothing. Pseudonymising, it receives bracket person one, bracket email one. With no filtering, it receives the real name and the real address.
14. Section five is the table a compliance officer takes to an audit: each article of the European regulation, what it requires, and what JMAX does about it. Pseudonymisation, minimisation, storage limits, right to erasure, protection by design and encryption.
15. And section six is the one that earns real trust: what the system does not do. That substitution is not magic, that no recogniser gets it right a hundred percent of the time, and that what an external provider does with what it receives is beyond its reach.
16. It also warns that encryption protects the disk, not the screen: with the application open and the session unlocked, the information is there for whoever sits down in front of it.
17. The last section tells you how to check all this yourself, with no special tools. Unplug the network and type a name and an ID number: you will see it still recognises them. Open a marked workspace and check the names are replaced.
18. In short: leave the firewall on, check that no internet provider sits on no filtering unless you decided so, and glance at the exceptions counter now and then. With that, no personal data goes out unless somebody chose it.