Most people store their files in one or two places: their phone’s camera roll, iCloud or Google Photos, and maybe a cloud drive. A photo from last night’s dinner sits next to a scanned copy of their passport. A casual voice memo lives in the same app as a recorded therapy session.
This isn’t a flaw in how people behave — it’s a gap in how storage apps are designed and marketed. “One place for everything” is a convenient pitch. It’s also a privacy model that treats all personal data as equivalently worth protecting, which means in practice that some things are overexposed and others are adequately protected by accident.
A better approach is tiering: deciding, deliberately, where different types of personal files live based on how sensitive they are and who could hurt you if they got out.
Why Tiering Matters
The answer isn’t “treat everything as maximally sensitive.” That approach leads to friction, overengineered solutions that people don’t actually use, and a false sense of security because real habits diverge from the ideal.
The useful question is: what’s the actual risk if this file is exposed?
For a photo of your lunch, the risk is essentially zero. For a document scan of your passport, the risk is identity theft. For a recording of a sensitive conversation, the risk is reputational harm or relationship damage. For a list of passwords written in a notes app, the risk is complete account compromise.
Treating these four files identically — because they’re all “personal files” — is a category error. They need different storage, different access controls, and different backup strategies.
The Five Tiers
Tier 0: Genuinely Public
Files you’ve already shared publicly, or content that doesn’t contain any personal information you’d object to anyone seeing.
Examples: profile photos you’ve posted on social media, screenshots of publicly accessible information, images you’ve published on a blog.
Storage: anywhere convenient. These files are already public by definition.
Key distinction: a photo you’ve shared on Instagram is Tier 0. The unedited original before you cropped and filtered it might be Tier 1 or higher, because it contains full-resolution details and EXIF metadata that the social media post doesn’t.
Tier 1: Personal but Low-Consequence
Files that are yours, that you’d prefer others didn’t have access to, but whose exposure would cause minor inconvenience rather than meaningful harm.
Examples: casual vacation photos, general notes about daily life, travel memories, screenshots of things you wanted to remember, grocery lists, public-event photos where you’re in the background.
Storage: mainstream cloud storage is reasonable here. Google Photos, iCloud, Dropbox, Amazon Photos — the tradeoffs (AI scanning, cloud-side key control, metadata collection) are real but proportionate to the low sensitivity of the content.
What to watch for: even low-sensitivity files accumulate metadata — location tags, timestamps, device information. The collection of this metadata over years creates a more detailed picture than any single file would suggest. Turning off location tagging on casual photos is a low-effort, worthwhile step for files in this tier.
Tier 2: Private and Personal
Files you wouldn’t share outside a small circle of trusted people, but whose exposure wouldn’t constitute a legal, financial, or physical risk.
Examples: personal journal entries, family photos, conversations with close friends, voice memos about your thoughts and feelings, private video recordings, photos of your home’s interior.
Storage: this is where the choice of provider starts to matter more. You want a service that:
- Doesn’t use your content to train AI models (yours or third-party)
- Has encryption in transit and at rest
- Doesn’t show your content in advertising systems
- Has a clear, enforceable deletion policy
Mainstream services that scan content for AI feature development are less appropriate here. Privacy-focused alternatives — services where the business model is subscriptions rather than data — are a better fit.
Practical note: personal journals and voice memos often contain information about other people — their private circumstances, your honest assessments of them. The sensitivity isn’t just to yourself; exposure could harm relationships.
Tier 3: Sensitive Personal Data
Files whose exposure could cause meaningful harm: financial loss, identity theft, reputational damage, or medical privacy violation.
Examples: medical records, test results, therapy notes, financial statements, tax documents, employment records, documents related to legal matters, personal correspondence about sensitive topics.
Storage: this tier demands more deliberate choices.
For cloud storage: choose a service with explicit commitments not to share data with advertisers or AI training services, clear encryption documentation, and verifiable deletion. Server-side AES-256 encryption (like daftei provides) ensures your files are encrypted both in transit and on the server — access requires authentication and is logged. This is meaningfully stronger than services that use your files for service improvement.
Consider whether cloud storage is necessary at all for some of these files. Encrypted local storage — a password-protected encrypted volume on your computer — eliminates the cloud-access risk entirely for files you only need to access from one device.
Access controls: enable two-factor authentication for any service where these files live. Use strong, unique passwords (managed by a password manager). Audit who has shared access.
Tier 4: Highest-Consequence Files
Files whose exposure could directly enable identity theft, account compromise, physical harm, or severe legal consequences.
Examples: government-issued ID scans (passport, driver’s license, Social Security card), financial account credentials, cryptocurrency seed phrases, private encryption keys, signed legal documents, anything that grants another person the ability to impersonate you or access your financial accounts.
Storage: these files should not be stored in mainstream cloud services.
Options appropriate for Tier 4:
- Encrypted local storage (VeraCrypt volumes, encrypted drives) on hardware you physically control
- Password manager vaults (for credentials specifically) — services like Bitwarden, 1Password, or Proton Pass are purpose-built for this type of sensitive data
- Physical storage for the most critical documents — a safe, a safety deposit box, printed backup of seed phrases in a sealed envelope
If you need occasional cloud access to Tier 4 files, use a zero-knowledge encrypted service where the provider cannot decrypt your content — Proton Drive, Tresorit, and similar services maintain this property by default.
The important distinction: daftei and similar privacy-focused personal storage services use server-side encryption (TLS 1.3 in transit, AES-256 at rest). This is strong security — access is encrypted and controlled — but it’s not the same as zero-knowledge encryption where even the provider cannot read your files. For Tier 4 content, zero-knowledge encryption is the appropriate bar.
How to Actually Apply This
The theory is straightforward; the implementation requires a one-time audit and then a habit of routing new files correctly.
The initial sort:
Go through your existing cloud storage and identify which tier most of your content falls into. You’re likely to find:
- The bulk of your files are Tier 1 — casual, low-consequence
- A meaningful subset are Tier 2 — personal but not catastrophic if exposed
- A smaller number are Tier 3 — sensitive enough to warrant more careful choices
- A handful are Tier 4 — these should be moved immediately if they’re in a mainstream cloud service
New file routing:
For each new file type that enters your life, decide once which tier it belongs to and where it goes. The goal is a habit, not a per-file deliberation. Once you’ve decided that medical records go to an encrypted local folder and casual photos go to your regular cloud, you don’t reconsider every individual file — you route it.
Metadata is part of the tier:
A low-sensitivity photo is higher sensitivity if it contains full GPS metadata placing you at a specific location. A public document is higher sensitivity if it has hidden metadata identifying the editor. Scrubbing metadata before sharing files doesn’t change their storage tier, but it changes what you expose when you do share.
What Happens When Files Change Tiers
Some files start at one tier and migrate over time.
A photo of a friend at a party might be Tier 1 when taken. If that friend is later involved in a legal matter or is a public figure, the same photo becomes more sensitive — potentially Tier 3 — because its exposure could affect someone beyond you.
A draft document you’re working on might be Tier 1 while you’re developing it. Once signed as a contract, it becomes Tier 3.
A casual journal entry from years ago might become more sensitive as your personal circumstances change.
Periodic review — annually is reasonable — is worth doing. The question isn’t just “is this where I want this file stored?” but “is this the right tier for this file now?”
The Practical Summary
| Tier | Examples | Appropriate Storage |
|---|---|---|
| 0 — Public | Posted profile photos, published content | Anywhere |
| 1 — Low-risk personal | Vacation photos, general notes | Mainstream cloud |
| 2 — Private personal | Journal, family photos, voice memos | Privacy-focused cloud |
| 3 — Sensitive | Medical, financial, legal docs | Encrypted cloud with strong commitments |
| 4 — Highest-consequence | ID scans, credentials, keys | Zero-knowledge or local encrypted only |
Most of your files belong in Tier 1 or Tier 2. The goal isn’t to make everything maximally secure — it’s to match protection to consequence, so the few files that could actually hurt you if exposed are actually protected.