We can download our data. But can we actually use it?

I recently caught myself asking an AI a question it could not really answer:

“What changed in my life over the last six months?”

The useful answer was scattered across my calendar, notes, photos, health apps, messages, purchases and location history. The model could help me think, but only after I collected the context and handed it over.

At that point, the privacy question became much bigger than “Which AI should I use?”

It became: Where does my personal history live, who controls it, and can I do anything useful with it?

Downloading is not the same as having control

In the EU, we have meaningful rights over personal data. Under the GDPR, people can request access to personal data held about them.

There is also a right to data portability in certain situations: when data is processed automatically on the basis of consent or a contract, a person can ask for data they provided in a structured, machine-readable format and, when technically feasible, have it sent to another provider.

That is slightly more precise than saying every app used in Europe must always provide a perfect “Download everything” button. The exact right depends on the data and how it is processed.

Still, the principle is important: personal data should not be a one-way contribution to a company’s database.

But anyone who has downloaded an archive knows the next problem.

You often receive a ZIP containing folders, JSON files, CSV tables and images with unfamiliar names. Technically, the data is yours. Practically, it may be very difficult to understand, combine or reuse.

So is that real control, or only compliance?

Our digital lives are split into pieces

One service knows our appointments. Another knows our exercise. Another has years of photos. Our notes contain decisions and half-formed ideas. Our messages hold relationships and memories.

Each service sees one small part of a person. The person is the only place where all those parts meet.

This becomes especially obvious when we use AI. People increasingly want to ask questions such as:

• What was I focused on before I burned out?
• Which habits actually improved my sleep?
• When did this project begin to drift?
• What promises did I make and forget?
• What stories from my family have I never written down?

A general model may be capable of helping with these questions, but it does not naturally have the context.

We can upload more personal information to another cloud service, but that creates another copy, another policy to read and another place to trust.

There should be another option.

What if the software came to the data?

Instead of sending our personal history to every new tool, we could keep a private, useful copy under our own control and let software work with it locally.

In that model, calendars, notes, photos and exported archives could be indexed on the device. Search could happen locally. AI could run on-device when the task and hardware allow it.

If a cloud model is needed, sharing could be optional, clearly explained and limited to the specific information required for that request.

This is not an argument that everything must stay offline forever. Cloud services are convenient, collaboration matters, and backups protect us from losing a device.

It is an argument for a different default: keep the personal source of truth close to the person, and make outside access a deliberate choice.

“On-device” is not a magic privacy label either. Local software still needs good encryption, careful permissions, reliable backups and understandable exports. A private system that loses all your memories with one broken phone is not a good system.

Maybe portability needs a more human definition

A useful personal-data tool should let us:

1. Retrieve our information without a battle.
2. Understand it without reverse-engineering dozens of files.
3. Combine information from different services into a meaningful picture.
4. Leave with an open, usable export whenever we choose.

That feels closer to ownership than a download button hidden in account settings.

I think the next privacy discussion should not be only about where companies store our data. It should also be about whether we, as individuals, can build value from our own information without constantly giving it away again.

I’m curious how other people see this:

• Have you ever downloaded your data from an app or platform? Did you manage to do anything useful with it?
• Would you prefer more software to work on-device, even if some features were slower or less convenient?
• What would a genuinely useful export look like to you?
• When AI uses personal context, should permission be given once, or separately for each topic and request?

This seems like one of those areas where the legal right exists, but the everyday experience has not caught up yet.

[Discussion](https://www.syndiary.com/news/we-can-download-our-data-but-can-we-actually-use-it/

Sources:
The European Commission explains the GDPR rights of access and data portability and the conditions that apply to portability requests.