There is a lot of talk about Gaia-X in the Italian newspapers, but obviously no journalist went to see what it is and how it works. Since one of my current clients is involved, who is one of the founders of the project, I will probably have to clash with it in the near future, so I had to start studying the existing documentation, which is also public.
First thing: whoever goes around saying that Gaia-X will save us from bad social networks or from the bad data management they do in the USA, has hardly read anything about the project. The need to have Gaia-X does not concern users of google or Facebook, but rather the service companies that already work on the continent.
Let's try to understand each other. Company X has an Industry 4.0 system that takes into account the supply chain, from suppliers to customers. It means that if customer X has bought a product, which I know is a television, I can know that it was assembled in that factory by guy, caio and semprionio, with pieces that came from companies A, B, C, lot D, serial number Z.
Well. The question is: where do I save all this flood of data?
And if you consider the amount of data of an industrial area like Europe, which is a mighty exporter, we are talking about REALLY LOTS of data. The truth is that all these companies, which form an impressive supply chain, will offer a VERY fragmented landscape. But if we REALLY want to work on the data, we have two possible solutions:
- in the chain of suppliers the “top” customer wins, the one who will ultimately shape the product (which I know, the Mercedes of the case), and all adopt the same technological solution as Mercedes.
- su uses an American cloud manager, all from the same provider, and at that point we have two or three technologies at play. But the data ends up in the USA.
- a gigantic moloch is created in Europe, and the first question is: who hunts the lira?
- federate existing services, building a federation layer between existing clouds in Europe. In short, everyone talks to everyone.
So what happens? Let's take an example with random names, so I'm not accused of advertising. It happens that the supplier, which I know Renault, has its own cloud and saves its data there, let's say in the Orange cloud. But to work as Industry 4.0 commands, it needs the data of its subcontractor, let's say Pippos Mechanicoi. The Greek subcontractor has the data in a cloud, which I know in the BasileaCloud cloud. If both are in the Gaia-X project, it happens that Pippos communicates to BasileaCloud his consent to share his data with Renault, through Orange. From that moment, Renault can perform a query on Pippos data, sending the query to Orange, which forwards it to BasileaCloud.
It is a matter of federating existing clouds.
This requires at least, I say at least, three things:
- that it is possible to identify Pippos and Renault with certainty, where "identify" indicates a robust and safe mechanism, in the telematic sense of the term. Infrastructure to build.
- that it is possible to identify with certainty, in the same “universe”, also BasileaCloud and Orange. To build.
- that there is a secure connection (for example a dedicated backbone) between BasileaCloud and Orange, which is difficult to intercept. Partially built infrastructure.
- that there is a "signature" capable of associating the data with its owner, or the one who can give consent to their treatment. That is, that the data can be identified with certainty. To build.
Having done this, we have created a European digital datalake where companies can (and probably will be obliged to do so with legal instruments) put data in a way that they can share while having control over it.
What is the advantage of this solution? Which does not require huge investments, and once implemented it can be used by any body on the European market .
It means that it is a solution oriented to the INTERNAL MARKET, to 100% European companies, and has almost nothing to do with the relationship between citizen and google, if not the fact that we want to prevent company data from ending up in the cloud of Google.
This obviously allows very convenient services for users as well: ideally if I have an object storage service on Magenta Cloud by Deutsche Telekom while you have it on SeeWeb, we can share files without needing me to make an account on SeeWeb or you on Deutsche. Telekom, which we should do instead, unless we create temporary download links, which lose track of consent and do not work as if it were a single cloud.
On the contrary, it is theoretically possible using Gaia-X for you to see my files in your bucket, and vice versa, which is already possible for two users of the same service from the same company. Think of Dropbox as an example: if you want to share data with another user, just give them permission.
This, therefore, allows the European market to have a layer of services that DO NOT depend on the USA, allows to overcome the differences between nations (not all nations have large clouds), but it does not have much to do with interaction. that passes between google and the common person: American companies have been deliberately excluded from having any decision-making power in the Gaia-X consortium, but after all, when you do a search on google you are talking to google, which will keep the data that you give your consent or not.
Ultimately, you can imagine a European dropbox where you can leave your data, but rather than being a single company it is the federation of many cloud companies. As in dropbox you can share data, make groups and everything, but then again, it is not a single company but a series of cloud providers working in a federated way.
Americans are afraid, but I don't speak (as the newspapers do) of the fear that Facebook has of not having your posts or your traffic on Whatsapp anymore. I'm talking about the fear he has of losing the business, because today companies are saving their data also on American clouds, which could be cut off from the business, simply because they don't federate .
Obviously, there is also the abstract possibility that having the power of every European cloud provider at its disposal, an Italian company could create the next Facebook, which today it could not do without having adequate infrastructures available (unless using US providers. , which would then eat it to stop the competition). So it is possible that a European whatsapp will be born, which is not possible today because there is no cloud in Europe that could handle attachments, voice files, etc.
But all this is theoretical, and it does not concern the problem of the user who has a Facebook account: it concerns the fact that banks and insurance companies can now share data (with the owner's permission!) For fintech services, which industries can exchange. data with their suppliers and customers, that the access permission can be revoked at any time, that if the data escapes it can be identified as the data that X managed on behalf of Y, etc.
Gaia-X is made for the European INTERNAL market, for businesses, for industries and for finance. It does not impact Americans if not by removing European customers from their clouds (especially if its use becomes mandatory) and has nothing to do with the privacy problem of the common citizen who uses Facebook, unless this infrastructure allows the birth of a Facebook in Europe.
I hope I have clarified, for everything else you can always read the specifications, which although they are of a high level you can find here: GAIAX
And it's time for the IT journalists to figure it out.
